CLASSIFICATION CHANGES - DTICaeroembolism or dysbarism). At 20,000 ft the incidence of symptoms is small and incapacitation is rare; raising the altitude above this point, however,

UNCLASSIFIED

AD NUMBERAD016860

CLASSIFICATION CHANGES

TO: unclassified

FROM: restricted

LIMITATION CHANGES

TO:Approved for public release, distributionunlimited

FROM:

AUTHORITYE.O. 10501, 5 Nov 1953; USAFSC ltr, 16 Oct1978

THIS PAGE IS UNCLASSIFIED

WADC TECHNICAL REPORT 53-227 SECURITY INFORMATION

DO NOT -L,"I)1yJ -3 RET!'• ' TO

p•,_f•_ i So •3 TECHýiA ,!ý " : ,','• ~CON1RO3. ui :li

AVIAuik com "BENDS" PAIN AS INFLUENCED BY ALTITUDEAND IN-FLIGHT DENITROGENATION

FRANKLIN M. HENRY

UNIVERSITY OF CALIFORNIA

MARCH 1953

WRIGHT AIR DEVELOPMENT CENTER

A&F-WP(9)-O-15 SEP 53 75 I~~

NOTICES

When Government drawings, specifications, or other data are usedfor any purpose other than in connection with a definitely related Govern-ment procurement operation, the United States Government thereby in-cur s no responsibility nor any obligation whatsoever; and the fact thatthe Government may have formulated, furnished, or in any way suppliedthe said drawings, specifications, or other data, is not to be regardedby implication or otherwise as in any manner licensing the holder orany other person or corporation,or conveying any rights or permissionto manufacture, use, or sell any patented invention that may in anywaybe related thereto.

The information furnished herewith is made available for studyuponthe understanding that the Government's proprietary interests inand relating thereto shall not be impaired. It is desired that the JudgeAdvocate (WCJ), Wright Air Development Center, Wright-PattersonAir Force Base, Ohio, be promptly notified of any apparent conflict be-tween the Government's proprietary interests and those of others.

This document contains information affecting the National defense of the UnitedStates within the meaning of the Espionage Laws, Title 18, U.S.C., Sections 793 and794. Its transmission or the revelation of its contents in any manner to an unauthorizedperson is prohibited by law.

RESTRICTEDWADC TECHNICAL REPORT 53-227 SECURITY INFORMA TION

AVIATORS "BENDS" PAIN AS INFLUENCED BY ALTITUDEAND IN-FLIGHT DENITROGENATION

Franklin M. Henry

University of California

March 1953

Aero Medical LaboratoryContract No. AF 18(600)-20

RDO No. 696-61

Wright Air Development CenterAir Research and Development Command

United States Air ForceWright-Patterson Air Force Base, Ohio

RESTRICTED

RESTRICTED

FOREWORD

Since a review of the existing literature revealed some seriousgaps in the basic knowledge necessary to cope with anticipated problems,and in order to secure addit~ional experimental facts on aeroembolism,a project identified as contract AF 18(600)-20, RDO No. 696-61, PO-12,entitled NHigh Altitude Physiology" was initiated. It was administeredunder the direction of the Aero Medical Laboratory, Wright Air DevelopmentCenter, with Dr. J. W. Wilson and Major David I. Mahoney acting as projectengineers.

The research work has been done by the Department of Physical Education,University of California, Berkeley, under the direction of the author.Acknowledgement should be made to the Donner Laboratory of Medical Physicsfor loan of the altitude chamber facility to the project, and to Dr. Johnf. Lawrence, M. D., and his Medical Physics staff for their helpful cooperation.Mention should also be made of Donald J. Rosenthal, M. D., who assumed the med-ical responsibility, Mr. Bruce M. Wilkin (a former B-24 pilot and oxygenofficer) who was in charge of the operations crew during the first half ofthe project, And the several un-named assistants who each played an importantpart in securing dependable data and in maintaining an excellent safetyrecord.

WADC TR 53-227

RESTRICTED

RESTRICTED

ABSTRACT

Experiments in the altitude chamber of the University of California(Berkeley) show that descent to a cabin altitude of 31,600 ft reduces the painof aviator's "bends" to moderate severity. Below 28,000 ft the pain is mild;it disappears (on the average) at 23,900 ft. Cases of bends kept at 10,000 or15,000 ft show a reduction in aeroembolism directly related to "storage" time;in.3 hrs there is a 13,000 ft gain in critical altitude. Frequently repeatedascent and descent does not change the critical altitude. With "storage" at20,000 or 25,000 ft the pain altitude changes very little within 3 hrs,presumably because there is growth of sub-clinical tissue bubbles. In-flightdenitrogenation for the prevention of aeroembolism symptoms should therefore becarried out at 10,000-15,000 ft. Oxygen economy can be aided by breathing cabinair at 10,000 ft during the first hour of a longer denitrogenation period onfull oxygen at 15,000 ft without sacrifice of protection. Available preoxygena-tion tables give valid predictions if physical exertion is mild; with heavy work,considerably more denitrogenation is required. Protection is achieved by reducingsymptom intensity rather than delaying onset of aeroembolism. Renitrogenationprogresses as the mirror image of denitrogenation.

The security classification of the title of this report is UNCLASSIFIED.

PUBLICATION REVIEW

This report has been reviewed and is approved.

FOR THE COMMANDER:

Colonel, USAF LMC)Chief, Aero Leadical LaboratoryDirectorate of Research

WADC 11R 53-227 iii

RESTRICTED

RESTRICTED

TABLE OF CONTENTS

Page

Introduction . . . . ... .

Section I Critical Pain Altitude . . .. . . . . 3

Section II Effect of Repeated Ascents and Descents . . . . 8

Section III Storage Experiments . . . . .9

Section IV Silent Bubble Experiments . . . . .. . 21

Section V In-Flight Denitrogenation . . . . . . . 22.

Section VI Renitrogenation Subsequent to In-Flight Denitrogenation 34

Summary . . . . . . . . .. . .. 37

Bibliography. . . 40

Appendix I Symptoms, other than Bends, Resulting from AltitudeChamber Decompression (Donald J. Rosenthal, M. D.) . 41

WADC TM 53-227 iv

RESTRICTED

RESTRICTED

AVIATORS "BENDS" PAIN AS INFLUENCED BY

ALTITUDE AND IN-FLIGHT DENITROGENATION

INTRODUCTION

When exposed to high cabin altitudes, some proportion of flight personnelwill develop symptoms of aviator's decompression sickness (also calledaeroembolism or dysbarism). At 20,000 ft the incidence of symptoms is small andincapacitation is rare; raising the altitude above this point, however, createsan increasingly grave problem. For example, at 38,000 ft, with physical activ-ity held to a low level, about 75% of individuals will develop bends or othersymptoms which will be severe enough to cause incapacitation in a third of thecases.if there is no preoxygenation. With a considerable amount of physical workgoing on, approximately 90% of individuals will develop symptoms which will beincapacitating in half to three-fourths of the cases.

The most common symptom is aviator's "bends", typically evidenced as adeep-seated pain in or near the knee, shoulder, or other joint. Less common butmore dangerous is "chokes", involving chest pain, difficulty in breathing, burn-ing sensation in the lungs, or perhaps uncontrollable coughing. Blurred vision,blind spots or other types of impairment may occur. There may be paralysis; theremay be gradual or sudden loss of consciousness, either independent of pain or asa result of severe bends pain.

Usually these symptoms disappear during descent to ground level, althoughthe severity of chokes frequently increases during descent and visual symptoms orshock may first appear after descent is completed. The consequences of aviator'saeroembolism have seldom caused permanent injury or death, but that is because thesymptoms are usually observed in an altitude chamber, where affected personnel canbe quickly restored to ground level pressures. Nevertheless, serious consequencesand occasional deaths from aeroembolism (dysbarism) are not unknown in the USAF (1).

The incidence of these various symptoms can be greatly reduced by severalmethods. Unfortunately, these methods are all subject to practical objections.Under normal modern flight conditions, symptoms are avoided by maintaining cabinpressure at a level sufficiently high to avoid aeroembolism. However, there arespecific flight situations of military interest when cabin pressure may be inten-tionally reduced or accidently lost. To cope with these situations, it isnecessary to consider and evaluate the available methods of prophylaxis.

One of these methods, preselection, will be only mentioned in passing, but itshould be kept in mind as definitely of potential value (3, P- 322; 8, p. 143) ifthe need becomes urgent, even though it is wasteful of personnel. It may be notedthat this method comes into use more or less naturally and inevitably in theroutine operation of an altitude chamber.

WADO Th 53-227

RESTRICTED

RESTRICTED

Another method, variously called preoxygenation, denitrogenation or oxygenpre-breathing, has been considered in a recent technical report (5), withparticular reference to certain flight patterns of current interest. In practicalcircumstances its use must be coordinated with the over-all engineering of theaircraft, since in-flight denitrogenation requires the breathing of pure oxygenand therefore the provision of an additional supply. It is not the purpose of thepresent study to examine that problem directly. It is intended, however, thatneeded background information will be developed in order to provide a basis forcalculating oxygen needs in projected long range bomber flights. In particular,the report cited above presented certain tables that were probably as accurate ascould be formulated from the available information. However, they involvedconsiderable extrapolation and are therefore in need of experimental validation atpoints of current interest. Moreover, the amount of oxygen necessary for someparticular amount of in-flight denitrogenation is dependent upon the altitude atwhich it occurs. While the higher altitudes are most economical with respect tooxygen supply, the decreased pressure is increasingly apt to produce rather thanprevent the symptoms that are of concern. Available data have not afforded anadequate basis for deciding the optimum altitude for in-flight denitrogenation.

A third method, recompression, has received very little experimental inves-tigation of a quantitative nature. It is important to know the minimum amount ofdescent that is necessary to secure partial or complete relief from aeroembolismthat has developed in flight. In addition, it is of interest to discover ifsymptoms that are temporarily relieved by partial recompression are still potentialand ready to recur immediately on re-ascent to a-higher altitude, or if they becomecured (either slowly or rapidly) during partial recompression. There is also needto determine just how much worse existing symptoms will become when the altitudeis increased by some specified increment, and to define the range of individualdifferences in this relationship.

In addition to the problems outlined above, several others should be mentioned.One that is closely related to the others has to do with the so-called "silentbubbles". According to theory, the symptoms of aeroembolism are due to the growthand expansion of bubbles formed from nitrogen and carbon dioxide released from thebody tissues. It is possible that exposure for a considerable time to an altitudethat does not produce clinically noticeable symptoms may nevertheless initiate sub-clinical bubble formation, so that ascent to a higher altitude will immediatelyproduce typical symptoms.

Another problem of concern stems from a few casual observations during theintensive aeroembolism research of the period 1942-45. With a subject in theairlock and in process of removal from the altitude chamber because of incapacitatingbends, an attempt was made to determine the critical pain altitude during descent.Having made one determination, the altitude was increased until the pain returnedto a moderately severe intensity and a second descent was made. This procedure wasrepeated several times. It was observed that the critical altitude became lowerwith each successive test, suggesting that rapidly repeated changes in altitudemight be greatly increasing the severity of the dysbarism attack. A few scatteredobservations on other individuals appeared to offer some confirmation of theeffect.

WADe TR 53-227 2

RESTRICTED

RESTRICTED

Quantitative information on and improved understanding of the various prob-lems listed above should be useful in several ways. The matter of planningoxygen supplies and flight profiles for combat aircraft has already been mentioned.In addition, it should be pointed out that information of this sort should proveimportant in estimating the need for and effective design of certain types ofemergency facilities, as for example pressurized suits or capsules. With theseproblems in mind, several series of experiments have been carried out and will bediscussed in the following pages.

SECTION I

CRITICAL PAIN ALTITUDE

Method

"Bends" pain, the most common symptom of aeroembolism, was rated in severityby the experimental subjects themselves using a large intensity scale postedinside the altitude chamber. (A reproduction of this scale is shown in Fig. 1).The pain rating technique as developed in this laboratory was validated andextensively used in decompression sickness research during the period 1942-45 (6).It should be noted that although the descriptive terms do not seem very sophisti-cated, they are the ones selected from a larger number of such phrases or terms,18 in all, on the basis of being most consistently ranked by an experimental groupof 20 subjects. The need for and. advantages of a multi-point objective pain scaleyielding reproducible results in bends research has been expounded elsewhere(3, pp. 332-334; 8, pp. 155-159). Other symptoms such as vasomotor reactions,chokes, and disturbed vision were diagnosed and evaluated by the project physician,who observed the subjects continuously during each experiment.

INTENSITY 5CALE 4 ALTITUDL5YMPTOM15o 1 Z 3456 789 CVIJe.. • a I i i i IU------

it % -s~ '

Figure 1. Pain rating scale.

W.0C TM 53- 227 3RESTRICTED

RESTRICTED

All subjects were healthy young male university students, 88% in the agerange 21-25 years and 98&' less than age 30. -They were paid $2.50 per hr foractual "flight" time. Prior to testing, each man was given a medical examinationby the project physician and an indoctrination flight to learn how to use theoxygen mask and to obtain practice in clearing the ears. At that time the generalnature of the experiment was explained briefly. BLB constant flow masks havinglight weight 2 liter rebreather bags were used, with adequate pressure to preventinboard leakage. The oxygen was moistened to reduce discomfort.

In the "storage" and "cycles"experiments which will be described presently,the subject was taken rapidly to 38,000 ft equivalent altitude in a decompressionchamber, where he performed 10 step-up exercises on a nine inch stool in 30 sec,at intervals of 2.5 min (California "D" test, 6). Upon the appearance of typicalbends pain of 40 intensity (which required an average time of 18.2 min, o- = 9.1),the altitude chamber was lowered at the rate of approximately 4,000 ft/min, withthe subject reporting by hand signal as the pain progressively lessened throughthe stages of 30, 20, 10, and eventually zero. Both the inside observer and thephysician observing from the outside were equipped with an altimeter placed withinthe field of vision, and noted the altitude (to the nearest 250 ft) at each intensityreport. This test was designated "original descent".

The subject was not permitted to see the altimeter during the test; he knewonly that the experiment would be concerned with varying altitudes less than40,000 ft. Having been cleared of pain, the subject immediately re-ascended towhatever altitude produced 40 pain and descended again, with the altimeter readingrecorded as before during both ascent and descent. The moment of reaching peakaltitude during this test was designated zero reference time. In the "cycles"experiment, additional re-ascents and descents were performed immediately, so that10 to 12 such tests were made during the course of an hour. In the "storage"experiments, there was an interval of some 20 to 25 min spent at a standard lowaltitude between successive tests, with the peak altitudes separated by 30 min.

Accuracy

The absolute accuracy of the various altitudes was controlled by using theaverage corrected readings of four new altimeters as a standard. Later on, a Haasstandard barometer was available for checking this figure. Considering also thevariability of the maintained chamber altitude, the average absolute accuracy isprobably dependable within * 85 ft at 10 to 20 thousand ft, and within ± 70 ft at30 to 40 thousand ft. It should be mentioned that since aeroembolism is the resultof a pressure differential, it was the practice to expose the men to a standarddifferential regardless of the ambient barometric pressure. However, the averageresults can be considered as absolute if the altitude of the Donner Laboratory(350 ft) is added to the reported altitudes.

WAD Th 53-227 4

RESTRICTED

RESTRICTED

Critical altitude for pre-established bends pain is considered to be theaverage of the altitudes at which pain of specified intensity appears during ascentand disappears during descent. Several errors are involved. There is evidence thatit requires a mean time of one or two minutes for the re-appearance of a symptomon re-ascent (9; 10), and we have observed that disappearance on descent is alsosluZggish. (A longer time at the test altitude would of course bring in the factorof bubble growth, creating an additional error). While the cause of the pain ispresumably the immediate re-expansion of extra-vascular tissue gas bubbles (3,pp. 33 and 198) since it recurs in the original site, some time is required forthis pain to develop and be perceived, and there is also some lag in rating andreporting. By averaging the ascent and descent data, these errors tend to cancelout. Other sources of error include inaccuracies in quickly reading the movingaltimeter pointer, and friction-induced irregularities in the pointer movement.In addition, the experimental subject's own reaction to the pain must be expectedto vary from one successive test to another, and it is probable that there are alsovariations of a physiological nature.

The over-all magnitude of these variable errors has been evaluated by deter-mining the test-retest correlation of individual critical altitudes, which turnsout to be r = 0.848 for 10 pain and r = 0.855 for 30 pain. In the case of 10 painthe standard deviation of individual altitudes is ar- 45 mm Hg in pressure units.Computing the standard error of individual measurements as -4T• - 17.5 mm Hg,and using the curve of Fig. 2 to convert to other units, it turns out that theerror is equivalent to 0.420 pain units or 1.35 thousand ft of altitude. For 30pain, a-- 43 mm Hg, hence the standard error of measurement is 16.4 mm (equivalentto 0.820 of pain or 1.5 thousand ft of altitude),

500

IdMILO I4ODERATC SEVERE -

I -25

ii SoE

300 /

/

0 /

// /1

,/ , /' /

/• F I

0 1 2 3e 4*

PAIN INTENSITY

rigare 2. Mean critical altitudeand fiducal limits for variousdegrees of pain intensity.

WADC TR 53-227 5

RESTRICTED

RESTRICTED

Psychological Controls

On seven occasions, two men who did the storage experiments simultaneouslyhappened to have almost identical critical altitudes. At the suggestion of theproject physician, several successive tests were conducted with these men sittingback-to-back. Care was taken to insure that they could not see each other's handsignals either directly or indirectly by shadows or by reflection from windows.The critical altitudes were just as similar when the men gave completely independentpain reports as they were when one man could see the other man's pain signals.

The California altitude chamber is equipped with two independent air inlets,one that is under the floor and silent in operation and another in the ceiling thatcauses air circulation and makes considerable noise when partly open. The vacuumline is exceptionally well silenced, and the pump is of relatively large capacity.These conditions made it possible to simulate descent while actually ascending orto simulate ascent while actually descending, insofar as external sensory cueswere concerned. On some 15 occasions "inverted" inlet noises were used for anextra test made as a continuation of the regular test. While some of the menindicated surprise to find the pain increasing again, there was no evidence that thecritical altitudes were influenced by the procedure.

On two occasions, when the project physician suspected that a subject was"faking" bends pain, a non-standard series of "cycles" tests was performed, usingthe inlet noises, slow ascent and descent, and showmanship on the part of thechamber operator, in an effort to completely confuse the subject as to altitudeand climb. The pain reports continued to be made consistently.

On two occasions during the seven hour tests that will be described in SectionV, the inside observer put on a pre-arranged "show" of the development of intensepain at the request of the project physician. Aided by appropriate comments overthe intercommunication system, these demonstrations were very realistic, and didapparently result in one man out of eleven reporting a temporary 10 pseudo-pain.

In the light of the above observations, both the project physician and thewriter are confident that the bends pains that constitute the basic data of thisreport are genuine pains, uninfluenced to any important degree by psychologicalfactors such as suggestion or pseudo-pains. Various informal psychological checksin addition to those mentioned have been made; quite uniformly, we have found thatour suspicions of invalid pain reports during the experiments have proved to bewithout foundation.

Experimental Results

Figure 2 gives the critical altitudes for different pain intensities for 56men who were tested 30 min after the original descent from 38,000 ft. The reasonfor choosing this particular time is8 that the critical altitudes are then lowest,as will be explainea below. Thirty-three of the men had two such tests, whichwere averaged to secure the individual mean scores. Since it was only intended tolet the pain develop to 40 intensity, the observation of 50 pain was uncommon andonly occurred in 12 cases. The average altitude increment for the °40-50 step wascomputed for these 12 men and added to the 40 average for the entire group, to givean adjusted mean for 50 pain.

WDC TR53-227 6

RESTRICTED

RESTRICTED

Critical altitudes for the other four pain intensities were determined byplotting the individual scores (ascent and descent averaged) as cumulativefrequency graphs on probability coordinate paper, and fitting a smooth curve tothe points. These curves were in turn used for determining the mean and standarddeviation to give the plotted points of Fig. 2 and the entries in Table 1. Thereis reason to believe, as explained below, that the 50 symptom represents apractical critical altitude in that greater intensities than this can be expectedto result in non-functional personnel.

Table 1

CRITICAL ALTITUDE OF PRE-ESTABLISHED BENDS PAIN

For various intensities of pain, in thousands of feet, as exhibited

by the specified percentage of individuals in the sample (n=56)

Inten- Cumulative per centsity -______

___y 10% 20% 30% 40% 50% 60% 70% 80% 90

10 19.7 21.2 22.2 23.0 23.9 24.7 25.8 26.8 28.6

20 22.6 24.0 25.2 -26.2 27.1 28.1 29.2 30.6 32.7

30 23.9 25.3 26.5 27.5 28.5 29.5 30.7 32.1 34.3

40 25.5 26.7 27.7 28.8 30.1 31.3 32.6 34.3 36.9

50 26.6 27.9 29.2 30.4 31.6 32.9 34.3 36.0 38.7

A conventional clinical interpretation of the pain intensity scores is shownat the top of Fig. 2. It is the opinion of the project physician that about halfof the observed 40 pains should be classed as severe so this has been taken asone of the transition points; similarly, at least half of the 20 symptoms areconsidered by him to be clinically moderate pains, so this is taken as the otherpoint of transition. It may be mentioned that in the 1942-45 research at thislaboratory, using the same pain rating system, the subjects remained at highaltitude until incapacitated (6). Pain within the 3--50 range was typicallyfound associated with impaired function, but only rarely with real incapacitation.More than half of the 60 symptoms were found to be associated with real incapacita-tion, so the 5--60 region can be considered another clinical reference point.

Conclusions fror the data will in general be based on analysis of 10 painreports and also 30 pain reports (as in our other publications), with the implicitassumption that it should thereby be possible to estimate the probable situationwith severe symptoms. In the present series of studies we were not permitted,

W&DC TR 53-227 7

RESTRICTED

RESTRICTED

for reasons of safety, to carry the symptoms to high intensities. It is thoughtthat the complete absence of severe post-flight reactions and hospitalized cases,in marked contrast to our 1942-45 experience, is due to this enforced experimentalcaution.

SECTION II

EFFECT OF REPEATED ASCENTS AND DESCENTS

"Cycles" Experiment

Immediately following the development of typical bends pain at 38,000 ft andtest zero, 11 men were individually subjected to the "cycles" experiment mentioned

previously. Variations in the individual critical altitudes for 10 pain as afunction of time are shown in Fig. 3. Points on these curves were obtained by

averaging the ascent and descent data for each cycle. Similar graphic analysishas also been done for 30 pain, with essentially the same pattern of results at analtitude about 4,500 ft higher. It may be seen that all of the subjects exhibit

"a lowering of the critical altitude during the early part of the period, reaching"a minimum followed by a rising phase indicative of symptom decay. On the average,the minimum is reached 26 min after test zero which corresponds.to 40 min after

symptom onset at 38,000 ft. The average of the minimum critical altitudes is 21.4thousand ft (cra 4.3).

N -24

/225 60

350 0 20

Is?3 // as

220 -as/

14 2440 0 to 40 1 It0

TIME (MIN)

Figure 3 Individual curves ofsymptom growth and dewa during cyclesexperiment.

2DC T 536-227

RESTRICTED

RESTRICTED

In order to compare these data with the critical altitudes obtained in theabsence of the cycling procedure, it is necessary to cross-section the "cycles"data at the time of the average minimum. At this point, the average criticalaltitude is 24.4 thousand ft, a figure which does not differ significantly from theaverage of 23.9 for 10 pain shown in Table I ("t" - 0.58). By a fortunatecoincidence, these latter data were secured at a corresponding time, namely 41min after symptom onset at 38,000 ft or 30 min after test zero.

The average altitude prevailing during the cycling procedure (i.e., theaverage of the highest and lowest altitudes reached during a cycle) was 28.0thousand ft at test zero and 26.7 thousand ft at the average minimum. Since thisis fairly close to the altitude at which one series of 21 subjects was "stored"for a period of 30 min following test zero, (25.7 thousand ft corrected altitude),it is perhaps more valid to compare the latter data with the cycles experiment.The average critical altitude for the test after 30 min storage was in this case24.7 thousand ft (c-= 4.1), which is again not significantly different from thecycles result ("t" -=0.17). It may therefore be concluded that the repeatedvariation up and down in altitude that characterizes the cycles test is withoutinfluence on the critical pain altitude.

The data of Fig. 3 are useful in illustrating the growth and decay process ofindividual symptoms. Considerable variation is apparent--in subjects Nos. 216, 220,225 and 229, for example, the major part of the growth had occurred before testzero. Some individual records, for instance Nos. 167 and 217, show a large andrapid growth and decay, whereas others, such as Nos. 60, 226, and 228 are charac-terized by a relatively flat curve. One case, No. 229, is of particular interestin illustrating that the curves of two different symptoms in the same individualmay vary quite independently. The average decrease in critical altitude from'testzero to the individual minimum points is 4.8 thousand ft, and at the average timeof minimum altitude, 1.8 thousand ft. These curves are very similar to symptomgrowth curves obtained by an entirely different method that have recently beenpublished in another report (5, P- 14).

SECTION III

STORAGE EXPERIMENTS

Method

In these experiments, typical bends pain was induced as in the "cycles"tests, and following test zero the subjects were "stored" at 10,000, 15,000, 20,000or 25,000 ft for 3 hrs. At intervals of 30 min, a quick test ascent to whateveraltitude produced 40 pain (or to 40,000 ft maximum), was made and followed byimmediate descent. In the case of the lower storage altitudes the rate of ascentant descent was 5 to 6 thousand ft/min up to the expected region of pain appearanceand 1.5 to 4 thousand ft/min thereafter.

The "ascent" and "descent" data for 10 and 30 pain were plotted separately onprobability graphs as shown in Fig. 4, using pressure ordinates rather than altitudeordinates since the former result in statistically normal frequency distributionswhereas use of the latter causes the distribution to be skewed.

VANRTR53-MIT 9RESTRICTED

RESTRICTED

20 -- 15,000 FT. STORAGE 01014.23

4.- a- -0

"" a

0' ' .? ~ ' , 91 2009L

400. .0

0 2350~

1 0 20 40 0 s 0 so 0 95 98 99

CUMULAT•o CASED S M

Figure 4. Individual data on changein pain from 30 to 20 intensity duringdescent.

In one of the examples shown, namely "original descent", it would have beenpossible to use the conventional numerical method for computing the mean and stand-

lard deviation since scores are available for each individual tested. In the otherexample, showing the results for the test made after three hours storage, it can beseen that the conventional method would give erroneous results. Half of theindividuals tested at this time did not develop 30 pain at the maximum test altitudeof 10,000 ft and therefore cannot be given numerical scores. It is possible, how-ever, to plot the points that are available, and by fitting a smooth curve, themean and standard deviation (a) can be determined graphically as shown.

In plotting these data the mid-frequencies have been used; e.g., the first,second and third individuals in the example have been assigned cumulative frequenciesof 2.2, 6.5, and 10.8% rather than 4.4, 8.7 and 13.0%. The mean is of course the50th percentile, while the standard deviation is the intercept at the 15.9 or 84.1percentile. Having determined the ascent and descent statistics separately, thetwo were averaged to give the critical altitude for 10 or 30 pain as the case mightbe.

W•GD TR 53-227 10

RESTRICTED

RESTRICTED

From 20 to 23 men were tested at each storage altitude. In a number of cases,the same individual was tested in more than one series, the amount of overlappingof this type being 33% between the 10,000 and 15,000 ft series and 45-501% betweenthe others. It may be seen in Fig. 5 that the groups were not perfectly matched ininitial critical altitude. However, the differences may safely be disregarded.They are no greater than would be expected from random sampling, since a varianceanalysis of the original descent scores yields an "F" of only 2.46 for 10 pain and2.26 for 30 pain, whereas an "F" of 2.75 would be required for statistical signi-ficance. The "F" coefficients at the 30 min test are also non-significant (1.27and 1.21).

too P

-30

STORlOEPIAINMIN

25 3-2,0 1

J IO I 25

• 0 '2 35

ý20 0 -0 AZ S

Figure 5. Symptom growth and decayat storage altitudes.

It should be mentioned that the mean critical altitudes for original descent($) were necessarily determined without averaging with ascent data. The pointsas plotted for this particular coordinate are adjusted means, obtained by usingthe 30 min test as a reference point and subtracting from this the pressureincrements between original descent and 30 min descent.

There was no test zero for many of the men in the 20,000 and 25,000 ft series(these were chronologically the first experiments). For this reason, a clearcut statistical evaluation of symptom growth during the first 30 min of storagemust be limited to a comparison of the change in "descent" critical altitudes only,using the differences between original descent and the 30 min descent.

WADO TR 53-227 11

RESTRICTED

RESTRICTED

Symptom Growth and Decay During Storage

Pooling data for the 15, 20, and 25 thousand ft storage series (since allthree show similar declines), the evidence is clear that symptom growth occursduring the first 30 min of storage at these altitudes. (The "t" ratio for altitudechange for 10 pain is 2.6 and for 30 pain is 5.5). In the case of the 15,000 ftstorage experiment treated separately, the decline is not statistically significantfor 10 pain ("t" - 1.4), but is definitely significant for 30 pain ("t" - 3-7).(For groups of the sizes used in this test, a "t" of 2.09 is required forstatistical significance). There is no evidence of symptom growth in the 10,000 ftstorage experiments.

Persistent or intermittent 10 and 20 pain was fairly common during storage at25,000 ft (33%) and 30 pain was sometimes observed (24%). Three of the men had tobe kept at substandard altitude during part of the storage, in order to keep thepain below 4o. (In two additional cases, not included in the statistics of Fig. 5,the pain was so severe that storage was impossible. Inclusion of these in Table 3would not alter the percentage figures). At 20,000 ft there were three cases of 10pain lasting for I to 1 hr, but no 20 pain. In the 15,000 ft series one individualhad 10 pain intermittently during.the first 30 min of storage. No pain was observedduring storage at 10,000 ft.

Table 2 gives the critical altitude range that includes 68% of cases, asdetermined by the six tests, for each of the different storage series. The twopoints for each test may be plotted at t loeon a probability graph such as shownin Fig. 4 and connected by a straight line, yielding a curve that permits the con-struction of detailed tables of the range of individual differences comparable toTable 1.

Turning now to the evidence for symptom decay during the main part of thestorage period, it may be seen in Fig. 5 that the rising trend indicative of this

. process is unmistakable for the two lower altitudes, and almost completely absentfor the two higher altitudes.

A statistical analysis of the 25,000 ft storage series shows no significantchange in critical altitude between the 30 min and 120 min tests for either 10or 3 pain. (This comparison is between the points of greatest difference).In contrast, the 20,000 ft data for both 10 and 3 pain show a statisticallysignificant rise in critical altitude at 60 min and every test thereafter, com-pared with the 30 min test as a reference point. While the rise is consistent,it is small in amount. Apparently the tendency for symptom decay after theinitial growth period is almost exactly balanced by continuing growth at 25,000ft, during the limited period of observation. Another and more attractive ex-planation would be that "silent" intravascular gas bubbles are interfering withdenitrogenation (2). This hypothesis, however, seems untenable because directmeasurements of nitrogen gas excretion show no change up to 35,000 ft and anincrease at higher altitudes (3, P. 310; 8, p. 56).

WAX TE 53-227 12

RESTRICTED

RESTRICTED

Table 2

FIDUCIAL LIMITS (+- C-) OF CRITICAL PAIN ALTITUDE

The tabled figures are in thousands of feet. Altitudeshigher than 40,000 ft were obtained by extrapolation.

Storage Storage time in minutes (30 pain or worse)Altitude 30 60 90 120 150 180

25,000 24.8 24.7 25.2 25.1 24.2 24.333.5 33.8 35.2 36.1 35.6 36.7

20,000 22.6 22.0 22.3 23.0 22.5 22.934.3 38.2 38.6 40.8 41.9 42.3

15,000 21.8 22.5 24.6 25.2 26.5 28.031.7 34.4 37.4 44.6 50.8 64.5

10,000 23.4 24.9 26.6 28.4 30.5 32.634.4 36.7 40.8 43.7 46.3 51.3

10,000 25.1 27.2 28.1 29.1 31.1 30.8(air) 35.2 38.6 40.8 41.3 42.8 43.7

Storage Storage time in minutes (10 pain or worse)Altitude 30 60 90 120 150 180

25,000 20.9 21.5 21.5 21.2 20.2 19.529.0 30.0 30.5 30.8 32.0 32.4

20,000 20.0 20.2 19-5 18.9 18.5 18.128.7 30.7 32.5 33.8 35.1 36.0

15,000 18.0 19.4 21.0 21.6 22.5 24.027.0 28.8 31.1 37-3 43.3 48.9

10,000 18.3 19.2 20.5 22.8 24.5 26.229.3 30.8 33.7 37.8 42.2 51.5

10,000 21.0 23.3 23.6 24.2 25.1 25.3(air) 29.2 32.5 35.3 36.1 35.8 37.0

WA TR 03-227 13

RESTRICTED

RESTRICTED

There is evidence from other stidies (5, P. 13) using a different technique,that the growth-decay curve shows a later and more gradual peak as physicalactivity is decreased and experimental altitude is lessened, so the presentresults are not entirely unexpected. It seems very likely that with a consider-ably longer period of observation, the 25,000 ft data would have exhibited asymptom decay phase evidenced by a rising critical altitude. As a matter offact, examination of individual curves reveals that 11 of the 21 individuals inthis series do show definite indications of symptom decay, although the averagingprocedure obscures this observation. (It will be recalled that the "cycles"experiment illustrated the wide differences in individual growth-decay curves).Evidently the region of 20-25 thousand ft is the altitude of slowest symptomdecay. At higher altitudes, there is an increasing apparent acceleration ofboth the growth and decay factors, while at lower altitudes the growth factorbegins to disappear, so that only the decay phase is operative. In this instancethe decay phase can occur only when there are pre-existing tissue bubblesinitiated at some higher altitude. At lower storage altitudes, the factor ofre-compression would be expected to assume an increasingly important role inadding to the symptom decay resulting from the denitrogenation that occurs atall altitudes (providing that alveolar nitrogen is absent or at low tension).

At 15,000 ft storage, there is still a symptom growth phase as mentionedearlier. It is becoming less evident than at higher altitudes and is apparentlynon-existent in the 10,000 ft experiments. The amount of re-compression is ofcourse greater at 10,000 ft. There is no alveolar nitrogen tension at eitherof these altitudes under the conditions of the experiment. It would be expected,therefore, that symptom decay would be greatest during the 10,000 ft storageprocedure. However, the data exhibited in Fig. 5 fail to agree with thishypothesis. There is no statistically significant difference between theresults at 10,000 ft and 15,000 ft for either 10 or 30 pain, analyzed as "gainin altitude" from original descent to the 180 min test or as gain in altitudecalculated from the low point at the 30 min test ("t t s" range from 0.8 to 1.1).In another analysis of these data, the individual critical altitude curves havebeen ranked in accord with the amount of individual symptom decay. A chi-squarecomparison of the results at the two altitudes shows no differentiation(I= 0.io4). Evidently the increased re-compression at 10,000 ft compared with15,000 ft is not important in producing any more rapid symptom decay. Thismatter will receive further discussion below.

Oxygen vs. Air During Storage

In order to study this situation further, an additional 10 000 ft storageexperiment was performed, with the men breathing ambient (cabinj air except ataltitudes above 12,000 ft during the brief high altitude tests. Fifteen of the21 subjects were tested in both experiments (71% overlapping), so the two serieswere very closely matched. (The critical altitudes at initial descent, and alsoat test zero, differ in the two series by only lY mm pressure). The comparativeresults of storage under the two conditions are shown in Fig. 6. At the 60 mintest, the critical altitude for the "air" storage appears to be somewhat higher,but this difference is within the limits of sampling error since "t" is only 1.3for 10 pain and 0.6 for 30 pain. After 3 hrs the critical altitude is lower inthe case of air breathing, and the difference is statistically significant("t's" - 2.4 and ?.3). The rate of symptom decay is therefore maintained afterthe first 60 min when oxygen is breathed during storage, whereas it declinessharply after this time when breathing ambient air.

WAD•C TR 53-227 14

RESTRICTED

RESTRICTED

so• Figure 6. Air vs. Oxygenduring storage.

0.000 Fr. STORAGE Oo.

o, o eoIto .. oSTOAGE TIME (MIN,

As a further statistical check, individual curves (averaging 10 and 30 data)for both series were ranked according to rate of symptom decay for the:first halfand separately for the second half of the storage period, and then analyzed bythe chi-square test. For the first half, the two series do not differ significantly(,X2 = 0.9). For the second half, the difference is definitely significant(x2 a 6.1). With oxygen breathing, there is no appreciable difference in therate of symptom decay during the first and second halves of the storage period(,X2 - 1.5). When ambient air is breathed, this difference is significant('X2 = 6.1); the rate of decay is very slow during the second half. The onlyintentional experimental variable is alveolar nitrogen tension, which is approxi-mately zero for oxygen breathing and about 400 mm when ambient air is breathed.In the latter case there is of course some hypoxia; the lips and fingernails ofseveral of the men were definitely cyanotic.

Considering the results at 15,000 ft and 10,000 ft breathing oxygen, theevidence forces the conclusion that during the first hour the rate of symptomdecay is independent of re-compression differences. At 10,000 ft, breathingpure oxygen causes no more symptom decay than breathing ambient (cabin) airduring the first 1 or l½- hrs even though it creates a greater nitrogen dif-ferential.

These findings can be reconciled with current theory, if it be postulatedthat both the 15,000 and 10,000 ft altitudes result in pressures greater thanthe critical amount for bubble growth (3, p. 202 ff.), which is quantitativelya more imrportant factor than the relatively small differences in resorbtion ratescaused by the ambient pressure differences. The other factor to be consideredis that the denitrogenation rate at 10,000 ft on air is faster than the "normal"rate. Jones (3, P. 310) as well as others (3, P. 142) have remarked on the factthat mild hypoxia produces a large increase in the denitrogenation rate. Itcan be seen in Fig. 6 that the trend of the data is in the direction predictedby this factor; the more rapid gain in critical altitude while breathing air duringthe first hour certainly suggests that this group is denitrogenating as fast or

•ter than the oxygen-breathing group. Evidently the greater nitrogen differ-e-"ttaj wihen breathing oxygen is more than counterbalanced by the greater rate ofgas exchange while on air at 10,000 ft during the early stage of denitrogenation.The effect is limited, of course; at 10,000 ft not more than 36% protection couldbe achieved with unlimited time. Assuming that the rate is increased to 30 minhalf-time (which is a reasonable assumption) 90% of this possible protection

WADC TR 53-227 15

RESTRICTED

RESTRICTED

would be reached in :1 hrs; this is just about the course of events shown inthe "air" curve of Fig. 6. After this length of time, the nitrogen differentialbecomes increasingly important in accordance with standard theory. Theexperimental results suggest that in-flight denitrogenation for the purpose ofbends prophylaxis might be as effective on ambient air as on oxygen during ap-proximately the first hour of the denitrogenation period.

It would seem reasonable to contend that a somewhat higher cabin altitudemight be desirable during the last half hour or so of the air-breathing denitro-genation period; both the increased hypoxia (provided it is not excessive) andthe greater nitrogen differential from the lowered air pressure would improvethe effectiveness. No doubt 12,000 ft could be used without running into dif-ficulty; possibly even 14,000 ft near the end of the period if the air crew wasacclimatized to some degree.

"True" Critical Altitude of Aeroembolism

To interpret the data properly, some allowance must be made for the factthat the effective storage altitude must have been greater than the base-line,because the tests made at 30 min intervals exposed the subjects briefly to muchhigher altitudes. By estimating the area of each individual flight profile itis possible to adjust the base-line to an average corrected altitude for eachseries, namely 141,200, 17,800, 21,500 and 25,400 ft. These corrections neces-sarily are only approximations, but they probably yield more accurate referencepoints than the nominal storage altitudes.

Using these figures, and avetaging the 10 and 30 data for maximum gain incritical altitude resulting from symptom decay during 3 hrs, the smooth curveof Fig. 7 has been constructed to define what might be termed the criticalaltitude for aeroembolism. (A comparable curve using the gain between test zeroand the 180 min test is very similar). It is probably a region rather than asharply demarked point, and may vary between individuals as well as with theexperimental conditions. The present study places the average critical regionbetween 18,000 and 21,000 ft, which is in excellent agreement with other studiesusing entirely different methods (3, pp. 217-218). It is somewhat lower thanthe mean critical altitude of 23,900 ft for pre-established bends pain reportedearlier in the present study, but this should occasion no surprise since thereis ample reason to believe that the physiological basis of decompression sick-ness exists at sub-clinical levels. The storage experiments, being concernedwith "silent" symptoms, prove this point very effectively. There was no clinicalevidence whatever of aeroembollsm at the lower storage altitudes while the menstayed at those altitudes, yet the evidence shows that the disease changed inseverity with the passage of time even though clinically "silent".

Results of Other Investigators

The results secured in a study by Fraser (2) on stepwise ascent may becompared in some respects with the current data. He used 1 hr at ground withoutoxygen or 1 hr at 10,000, 20,000 or 27,500 ft with oxygen, followed in each caseby 2 hrs at 35,000 ft. Using the non-denitrogenated ground level group as acontrol, the total symptom incidence was reduced to 45% of the standard in the

WADC TR 53-227 16

RESTRICTED

RESTRICTED

10,000 ft series; 65% in the 20,000 ft series, and 70% in the 27,500 ft series.Considering only-moderate or severe symptoms, the figures were 27, 44, and 70%of the control incidence.

Fraser attributed the higher incidence with the 20,000 and 27,500 ft -steps"to the formation of silent bubbles that interfered with the denitrogenationprocess. Whether this hypothesis is accepted or not, the results do show with-out question that in-flight denitrogenation is less effective at 20,000 ft orhigher than is the case at 10,000 ft. The relative incidence of total symptomswith the 10,000 ft step is very close to the prediction of 55% obtained from theJones tables (3, P- 318; 8, p. 98); moderate and severe symptoms have beenreduced more than the average expectancy for one hour of ground level denitro-genation. Gray (4) observed that j hr of denitrogenation at 14,250 ft was aseffective as at ground level, whereas at 19,250 ft or higher it was less effective.These studies agree in indicating that in-flight denitrogenation at 10,000 andprobably 15,000 ft is no less effective than at ground level.

CORRECTED STORAGE ALTITUDE (M FT)15 20 25I5C I I I I I I I

S15

0

00

tI-Z

M o

IL

a 0 ;ý 5 0

x0

a-

o0 I PAiN

X.?3 PAiN

400 :350 300

CORRECTED STORAGE ALTITUDE (MM HO)

Figure 7. Symptom decay related to storage altitude.

WADC TM 53-227 17

RESTRICTED

RESTRICTED

Individual Differences

Some interesting side-lights of the storage experiments justify additionalcomment. One such matter has to do with the question of whether individuals whoexhibit a low critical altitude during the early part Of storage show more orless symptom decay than those who have a relatively high critical altitude. Usingthe 15,000 ft storage data, the 12 "lowest" and 12 "h~ghest" individuals werecompared for amount of symptom decay. For 10 pain, IX was 0.4, and for 30 painit was 2.1. Neither of these statistics is significant. If there is any dif-ference, it is apparently not of much importance.

There are fairly consistent individual differences in the rate of symptomdecay at the lower storage altitudes. Pooling the 10,000 ft and 15,000 ft data,there are 26 cases who had storage tests on two separate occasions. For eachseries, individual decay curves have been ranked as above or below average forthat series. It turns out that 77% of these men hold their relative positionsas having a "fast" or "slow" curve in both tests. A chi-square analysis gives,x2 = 6.4, which is highly significant statistically, and corresponds to a correla-tion coefficient of r = 0.74. Presumably these differences are due to individualdifferences in rate of nitrogen clearance.

The reproducibility of individual critical altitudes from one day to anotheris somewhat less than this, namely r a 0.73 for 10 pain and r = 0.73 for 30 painduring original descent from 38,000 ft. As might be expected, this day-to-dayreliability is somewhat less than is the case for two successive tests made onthe same day. Earlier in the report, the reliability of the latter was found tobe r = 0.85 for 10 pain and r = 0.86 for 30 pain.

Percent Recurrence After Storage

It is possible to compare certain of the data secured in the present

experiments with results obtained by others, although it is necessary to limitthe comparison to relatively crude descriptive statistics. Stewart and Smithin 1943 (10) reported that the recurrence of pre-established decompression

sickness, tested by re-ascent after storage at ground level for J to 1 hrbreathing air, was 100%. This is in sharp disagreement with results reportedby Rodbard in 1944 (9); the latter investigator found only 50% recurrence.With 1½ to 2 hrs storage, the Stewart and Smith finding is 75% recurrence ofpain in the original site compared with 52% reported by Rodbard; with 2i to 3hrs storage the figures are 77% and 38%. The Rodbard data show a statisticallysignificant difference from the others in each of these comparisons. A possibleexplanation of the discrepancy is that the method and criteria for bends weredifferent in the Rodbard study. Reference was made earlier to the necessity fora standard methodology in bends research. In the present investigation, 10,000ft storage breathing ambient air (with, of courfe, less re-compression) resultedin 98% recurrence after 2 to 1 hr, 74% after 11 to 2 hra, and 67% after 2t to 3hrs storage.

WADC Ti 53-227 3.

RESTRICTED

RESTRICTED

Stewart and Smith also made a few observations on storage at several higheraltitudes. Four men were stored at 25,000 ft for 30 min with complete recurrenceand 16 men were stored at 20,000 ft for periods running from 30 to 120 min withcomplete recurrence. Without question, these findings agree with the presentstudy (Table 3). Nine men were stored by them at 15,000 ft for 30 min, withonly 44% recurrence--much lower than the amount of recurrence at the same altitudein the present experiment or in their storage at ground level or at higheraltitudes. Due to the small number of individuals, which results in a largesampling error, it is doubtful if much significance should be attached to their15,000 ft experiment. In Section VI of the present report it will be shown thatthe proportion of recurrence as a function of time, in the case of both 10,000and 15,000 ft storage on oxygen as reported in Table 3, decreases in accordancewith denitrogenation theory.

Table 3

PER CENT RECURRENCE OF ESTABLISHED BENDS PAIN

Storage Storage time in minutes (30 pain or worse)Altitude 0 30 60 90 120 150 180

25,000 100 95 95 95 95 91 91

20,000 100 90 85 85 80 80 80

15,000 100 100 96 87 74 61 52

10,000 100 100 91 81 57 43 38

10,000 100 100 95 76 71 67 67(air) I

Storage Storage time in minutes (10 pain or worse)Altitude 0 30 60 90 120 150 180

25,000 100 100 100 100 95 100 95

20,000 100 100 100 90 90 85 85

15,000 100 100 100 100 91 83 65

10,000 100 100 100 91 86 67 57

10,000 100 100 100 100 95 86 86(air) _

WADC TR 53-227 19

RESTRICTED

RESTRICTED

Comment on Reproducibility

The "Califorhia D" test, used to initiate bends pain in the current storageexperiments, has in other investigations been reported to produce an incidenceof 87% total symgtoms, usually within the first 45 min, with 90% of the casesprogressing to 3 or worse (3, p. 231; 6, pp. 42 and 48). In the unselectedgroup used in the present experiments, 89 men had an original 38,000 ft "D"exposure of 30 min or somewhat longer to pre-establish the bends studied in thestorage or cycles series. The total incidence was 89% with 84% of the casesprogressing to 30 or worse. (A similar correspondence is evident between thelight exercise series to be discussed in Section V and an earlier Californiastudy). The remarkably close agreement between the incidence in these twostudies (separated in time-by about ten years) is mentioned here in order toestablish that the subjects of the present study constitute a representativegroup, and to emphasize that it is possible to secure reproducible results inaeroembolism research if meticulous attention is given to the standardizationof methodology. Lack of such standardization has greatly limited the usefulnessof experimental work (otherwise sound) done in different laboratories (e.g. 5,p. 21. Also see 3, pp. 225 and 332 ff. or 8, p. 154 ff.).

The primary datum in human aeroembolism is the pain experienced by theperson himself. This is, of course, subjective, but is adequately consistentif proper psychological conditions are maintained. It is not consistent orquantifiable if they are not maintained.

If the men undergoing the experiment are oriented to hide their pain,either by direct instruction or by the general atmosphere of the procedure,some individuals will do so and some will not, hence the resulting data willinclude an additional (although avoidable) source of error variance as well assystematic error. If the orientation is intentionally or unintentionally inthe direction of over-emphasizing the severity of symptoms or the danger in-volved, (as may well occur for example under poor psychological conditions orin improperly indoctrinated groups) there will be an additional avoidable sourceof error variance and systematic error. Data biased by these errors are verydifficult to interpret; one can only guess at the quantitative influences ofthe biasing factors.

Beyond the primary datum, it is possible to maintain objectivity, but thiscan be accomplished only if the subjective influence of the experimenter isrigorously excluded. There is, unfortunately, a constant temptation for theexperimenter to decide how severe the pain must have felt to the subject, butany such decision constitutess'a secondary datum involving two subjectiveelements. These simple straightforward facts cannot be avoided by limitingthe reports to severe pain. Such a procedure may even make the situation worsefrom the point of view of securing reproducible data. The experimenter canmeasure rather accurately (although not very practically) how many men loseconsciousness in an aeroembolism experiment. However, he cannot know if a manwho is "putting on a show" is suffering either more'e or less acutely than someother individual who quietly signals a pain of the same intensity, or is closerto or further from complete collapse, unless he secures evidence to validatehis opinion, and keeps record of his failures as well as his'successes. (Sucha record is illuminating, although not very flattering).

WAZC TR 53-227 20

RESTRICTED

RESTRICTED

Evidence is available to validate the above-mentioned primary data in apractical frame of reference. Motor impairment, rated during muscular activityby an experienced observer without knowledge of the degree of pain reported bythe man suffering from bends, has been found to correlate r = 0.90 with the painas evaluated by the method used in the present investigation (6, p. 37).

SECTION IV

SILENT BUBBLE EXPERIMENTS

Formation at 15,000 Feet

The experiments previously described have furnished ample evidence of the

existence of "silent" extra-vascular tissue bubbles present at storage altitudesas, residual from an immediately preceding higher altitude exposure (also see 3,pp. 33, 193 and 197 ff.). Some additional data have been obtained under condi-

tions where the initial bubbles are established at the storage altitude. Sincethe other experiments clearly indicated bubble growth at 21,500 ft and 17,800 ft,

with no growth at 14,200 ft (corrected altitudes) it was decided to use a storagealtitude of 15,000 ft. Interest had also been directed to this particular altitudeas the highest that could safely be used for in-flight denitrogenation purposes.

There was no pre-storage exposure to higher altitude, and the "tests" werereduced in number and spaced further apart so that the nominal storage altitudewould be very nearly the same as the effective altitude.

With indications that there would otherwise probably be no silent bubbles,

it seemed desirable to use exercise as a facilitating agent to start bubbleproduction. Twenty-one men exercised at 21 min intervals for the first J or 3/4hr of exposure (using the standard 10 step-ups mentioned in connection with othereiperiments) and 11 men exercised in the same manner but at 5 min intervals forlhrs.

The last-mentioned group of 11 men had no bends symptoms when tested with a

quick test climb to 40,000 ft after exposure times of lA, 3 and 4 hrs, with theexception of one man who'had a 10 knee pain that came in at 38,000 ft or, the firsttwo of these tests.

Among the men who exercised more frequently but for a shorter time, five

exercised for 2 hr and were given the high altitude test at half-hourly inter-vals for 4 hrs, with no symptoms. Ten exercised 3/4 hr and were given the testat hourly intervals beginning immediately thereafter (4 men) or after 15 minrecovery from t e exercise (6 men). Wo cases of bends pain occurred, but one

man exhibited 1 abdominal gas pain at 1 hr, 30 at 2 hrs, 40 at 3 hrs and 5 0 at

4 hrs. The remaining six men were given only two high-altitude tests--immediatelyafter exercise, and at the end of 2 hrs. Three had no symptoms, one showed a 10

ankle pain at 39,000 ft during the first test and the other two had more severepains. One of these cases reported an ankle pain of 10 at 26,250 ft and 20 at31,000 ft on the first test that was absent on the second test. He also had 30

abdominal gas pain at 40,000 ft on the first test and 20 at 40,000 ft on the

second test. The other case had 10 knee and ankle pains on the first test, thecritical altitude being 33,000 ft. On the second test the ankle pain was absent,but the knee pain responded at the following critical altitudes: 10 at 27,000 ft,20 at 29,250 ft, 30 at 31,250 ft, 40 at 33,000 ft, and 50 at 35,250 ft.

WADC TR 53-227 21

RESTRICTED

RESTRICTED

These results may be summarized as follows: In 32 men showing no symptomswhatever during storage at 15,000 ft, the high altitude tests showed mild bendsthat receded in two cases (6.3%) and severe bends developing progressively though"silently" during storage in one case (3.1%). Abdominal gas occurred in the sameproportion.

SECTION V

IN-FLIGHT DENITROGENATION

Heavy Exercise Series

Twenty of the men in the series described in the preceding section werekept at 38,000 ft from 11 to 2 hrs after the silent bubble experiment, in orderto test the effectiveness of the 15,000 ft storage as an in-flight denitrogena-tion procedure. In four of the men, (the ones who had somewhat less than 4 hrsof in-flight denitrogenation), three showed 10 pain appearing within 12 - 40 minafter reaching 38,000 ft and regressing later. The other case also had bends,beginning with 10 after having beenat the high altitude 28 min, and reaching 40

intensity 18 min later. He was then removed from the altitude chamber. Thesefour men did the step-up exercise at 2' min intervals for the first 45 min at38,000 ft.

The other 16 men, with a full 4 hrs of in-flight denitrogenation and exerciseat 15,000 ft, went to 38,000 ft for 2 hrs with exercise as above. Six weresgmptom-free, seven had bends that receded (five went to 10 only, one went to2 and one to 30). Two men were forced to descend, one with bends that reached40 after 24 min exposure, and one with chokes at 26 min. One man (not includedin the statistics) went only to 35,000 ft because of abdominal gas pain. Hedeveloped no bends, but had to descend after only 44 min of exposure.

Eight additional men had no "tests" or exercise during the 4 hrs of in-flight denitrogenation, and exercised only during the second hour at 38,000 ft.Six were symptom-free. Two men developed serious symptoms after the exercisebegan. One man had bends pain that reached 40, causing descent at 1 hr 33 min(i.e. 33 min after he started exercise). The other man had 40 bends and a vaso-motor reaction (VMR) indicative of impending collapse at 1 hr 54 min.

Summarizing these data, and correlating with the preceding section, thefollowing implications may be derived: Silent bubbles can originate at 15,000ft when there is exercise. In most cases they are small in magnitude and recededuring 3 or 4 hrs. If there is heavy exercise at this altitude, there will bean occasional case (3%) of profuse sub-clinical bends that will cause immediatetrouble upon ascent to higher altitude. While the in-flight denitrogenationthat occurs while breathing pure oxygen at 15,000 ft exerts a considerableprotective influence, it is inadequate for reasonable protection against exposureto 38,000 ft if there is heavy exercise at this higher altitude. In this situa-tion, moderately severe bends or chokes can be expected in 18.5% ± 7.6% cases,since five cases of this type were observed in 27 men. The total incidence ofsymptoms was 56%. Controls were obtained on 13 of the men, who repeated their

WLDC M 53-2.7 22

RESTRICTED

RESTRICTED

38,000 ft schedule without any type uf oxygpn pre-breathing. Forty-six per centdescended with chokes, 40 bends or 50 bends; 54% had 30 bends or worse; the totalsymptom incidence was 859. These figures cannot be used as a fair 4 uantitativetest of the validity of preoxygenation tables previously reported (5, P. 19),which predict a much larger effect from the denitrogenation. No more thanmoderate physical activity is presumed by those tables. Furthermore, the numberof individuals in the sample is rather small. Regardless of qualifications, theobservation of serious symptoms in nearly a fifth of 27 men tested is adequategrounds to deny that the denitrogenation afforded satisfactory protection.

Cause of Poor Protection

There are indications within the data of the present study that the inadequacyof the protection is chiefly due to tlie exercise at the higher altitude, inasmuchas the time of the serious symptoms is apparently related to the time that theheavy exercise occurs. The nitrogen component of the gas bubbles that presumablycause the pain or other symptoms is of course reduced by breathing pure oxygen,either at ground level or in flight at 15,000 ft. On the other hand, thisprocedure can have no influence on the local concentration of carbon dioxide,which is probably the main causal agent that is responsible for exercise increas-ing the incidence and severity of aeroembolism (3, pp. 232-233). There hashowever been some controversy concerning the importance of carbon dioxide--Ferris(3, P. 33) dismisses the systematic experimental evidence (3, p. 232) while Cook,(3, p. 230) questions the interpretation that Ferris has placed on his own data,concluding that it actually supports the carbon dioxide hypothesis. The "exerciselimit" emphasized by Jones in his analysis of preoxygenation data (3, PP- 304-306;8, p. 51) is very likely a reflection of the importance of carbon dioxide.Animal experiments with preoxygenation also demonstrate the role of carbondioxide in bubble formation (3, P- 150 and 154).

The above line of reasoning would seem to lead to the hypothesis that therelative influence of exercise on aeroembolism should be markedly increasedunder denitrogenated conditions. While this particular idea has never beentested by systematic experimentation on humans, there is certainly ample evidencethat dependable protection for exercised exposure gained by preoxygenation atground level often requires far more than four hours of denitrogenation.Unfortunately, not all the available data on preoxygenation can be accepted atface value (3, p. 253 ff.).

If, for example, an individual is repeatedly exposed to aeroembolism alti-tudes in a series of experiments involving progressively longer periods of pre-oxygenation, until in one of these experiments he fails to have symptoms, it isquite wrong to naively conclude that this amount of denitrogenation has beenshown to "protect" that individual. On the contrary, there is a fairly highand known probability that the absence of symptoms is simply a reflection ofintra-individual variability (3, pp. 325-330; 8, pp. 147-153). In two flightsunder the same conditions, results from five different laboratories have shownthat on the average about 40" of men forced to descend in one flight remain upin the other, and about 18% of men who remain up in one flight are forced todescend in the other. Figures on symptoms vs. no symptoms are similar to theabove. (Assuming constant reproducibility, such percentage figures will varydepending on the incidence in the particular group) (3, P. 336 ff.; 8, p. 161).

WAD0 T% 53-227 23

RESTRICTED

RESTRICTED

For this reason, definition of the amount of preoxygenation required forthe protection of a particular individual requires that he be repeatedly testedat that particular dosage. Similarly, the protection afforded a group ofindividuals cannot be estimated unless the data of all members of the group areincluded, quite regardless of whether or not certain of the individuals in thegroup may have seemed to be "protected" by a lesser amount of preoxygenation.Interpretation of published data on the amount of preoxygenation required forprotection must be conditioned by these harsh but realistic facts; dependableprotection is more difficult to achieve than is commonly realized.

Light Exercise Series

It has been emphasized in another report that insofar as predictions canbe made from available data, four to six hours of denitrogenation will affordadequate protection "provided that muscular activity be limited to mild and in-frequent exertion" (5, P- 1). This is admittedly a rather un-precise statement.Estimations-of the amount of activity ord4narily engaged in by flight personnelin the military situations where aeroembolism is anticipated are also difficultto quantify. Wide differences in opinion have been voiced. Having intervieweda number of B-36 personnel and inspected the various positions in this particularaircraft, the writer is of the opinion that the amount and type of activityunder the conditions of interest should be defined as light activity, probablycomparable on the average to card-playing with occasional periods of standingup and stretching. The writer is unable to state a dogmatic opinion as to howtypical this is of the activity in other aircraft of current interest. Ifpersonnel are wearing pressurized T-1 suits, it is fairly certain that heavyexercise will be uncommon. Under emergency conditions, it may sometimes happenthat a particular crew member will find it necessary to engage in heavy exercise.If so, his probability of having severe aeroembolism will thereby be greatlyincreased. However, it is not certain that even 8 or 10 hours of denitrogena-tion will guarantee protection in this circumstance (although the probabilityof trouble would be lessened).

In the light of the above considerations, and because of other reasons forinterest, data on a standardized series of in-flight denitrogenation tests withlight exercise have been secured. The activity was card-playing as describedin the previous paragraph. Two flight profiles have been used. In one, themen breathed ambient (cabin) air for 1 hr at 10,000 ft followed by 3 hrs onundiluted oxygen at 15,000 ft and 3 hrs high altitude exposure at 38,000 ft.Reasoning from the data shown in Figs. 5 and 6, it seemed likely that thereshould be very little difference between the effectiveness of this profileand a "standard" profile using full oxygen at 15,000 ft during the first houras well as the subsequent 3 hps before the high altitude exposure. Shouldthe experimentalfacts support this hypothesis, it would be possible to comecloser to achieving the recommended amount of denitrogenation (5, p. 2) inoperational flights, while at the same time conserving oxygen and lesseningthe amount of discomfort to personnel that is unavoidable with the recommendedspecial oxygen discipline.

It was also thought desirable to secure a control series, using the sameconditions of exposure at 38,000 ft, but reducing denitrogenation to a minimumby starting the oxygen supply at 12,000 ft on the way up, with a fast ascent.It was realized that the control series would not represent ideal data, sincewe were not permitted intentionally to carry the men to the point of true in-

WAfC TE 53-227 24

RESTRICTED

RESTRICTED

capacitation. Nevertheless, it would serve to give some indication as to whatwould happen without the denitrogenation, and could tie in the present datawith experimental work done in this laboratory some years ago as well as datafrom other sources.

A total of 29 men were tested on the 10,000-15,000-38,000 ft profile and29 on the 15,000-15,000-38,000 ft profile, with 52% of overlapping individualsin the two groups. In other words, in the two combined there were 58 tests on43 individuals. The control series consisted of 40 individuals, overlapping."90% with the first-named denitrogenation group and 80% with the second orstandard group. Of the controls, 62% of individuals did the control after hav-ing done a denitrogenation test. Nearly every person used in any of theseseries had previously seen bends in one of the other experiments describedearlier and a majority of them had previously experienced and reported thepain themselves at one time or another. The individuals are thought to consti-tute an unselected sample.

Experimental Results

The results are shown in Table 4. There was no case of true incapacitationin either of the !n-flight denitrogenation groups, although there is a possibilitythat the individual who was ordered down after 1 hr 28 min because his painreached 50 in the 10,000-15,000-38,000 ft test might have become incapacitatedif he had remained at 38,000 ft. There is also a good possibility that he mightnot have become incapacitated. It really makes very little difference just howthis individual is classified. It is difficult to define incapacitation in theabsence of syncope or paralysis; moreover an occasional case of serious bends issure to occur with even considerably more denitrogenation than used in thesetests (3, p. 253).

It should also be mentioned that one individual in this group (who had a 20bends pain at 1 hr 33 min which thereafter regressed) became nauseated justeight minutes before the end of the 38,000 ft exposure and was taken out at oncefor reasons of safety. His blood pressure was normal and he had no symptoms ofthe type usually associated with impending vasomotor collapse, except sweating.It was his own opinion, with concurrence by the physician, that the nauseawas due to the fact that he had been up most of the previous night, and,contrary to instruction, had eaten no breakfast before the beginning of the7 hr altitude chamber flight. His nausea persisted for about 1 hr afterdescent before any substantial improvement occurred. Otherwise, recovery wasuneventful. Uncomplicated, late appearing nausea is not a common symptom ofaeroembolism; this man's symptoms may or may not have been related to the highaltitude exposure.

All subjects in the 15,000-15,000-38,000 ft group completed the full 3 hrsat 38,000 ft. One case had a bends pain that developed to 40 after 51 min athigh altitude; this pain regressed to 30 but did not disappear. This individualhad discomfort, but showed no evidence of incapacitation. On the whole, thesymptoms tended to be slightly less severe in this group, but this is simply arandom variation or "sampling error" because the control tests of individualsin the group also show somewhat fewer and less severe symptoms than men in theother group; descents due to 40 or 50 bends, or to chokes or VMR, was 44% com-

WADe TR 53-227 25

RESTRICTED

RESTRICTED

pared with 59%. While this difference is not statistically significant(X2 = 0.69), the fact remains that the individuals were on the average lesssusceptible to severe aeroembolism than in the other group; the effectivenessof the denitrogenation was no greater.

Effectiveness of First Hour on Air

Compared on the basis of the incidence of 20 pain or worse, the two setsof controls are perfectly matched at 52%, and the two denitrogenated groups arealso perfectly matched at 17%. Compared on the basis of total symptoms, the"air" denitrogenation group had fewer symptoms; on the basis of 30 or worse ithad more symptoms. In neither case is the difference statistically significant,even though it be assumed that the intercorrelation was as high as r = 0.60,which is a too-generous allowance (3, P- 341). Compared on the basis of 40symptoms or worse, there is no difference. It is therefore necessary to concludethat the first hour of a 4 hr-in-flight denitrogenation procedure is practicallyas effective utilizing ambient air breathing at 10,000 ft as using pure oxygenat 15,000 ft, even though the alveolar nitrogen differential is somewhat less(during that hour). There is every reason to believe that this finding wouldhold true if the oxygen breathing were extended to a longer period; it may ormay not be true for shorter periods.

It cannot be claimed that these results have definitely proved that thaireis no difference whatever in the effectiveness of the two profiles. All thatcan be concluded is that any such difference must be very small and unimportantpractically, as otherwise it would have become manifest in the comparison ofthe two well-matched groups.

There is now justification for combining the two series in order to secure amore reliable prediction of the results to be expected with 4 hrs of in-flightdenitrogenation. To do this, it is desirable to first average for each individualthe results of his two tests (if he had two); some individuals were tested inboth series and some in only one. The size of the sample is 43 men.

WA&De T 53-227 26

RESTRICTED

RESTRICTED

Table 4

PROTECTION FROM DECOMPRESSION SICKNTESS

Secured from two profiles of in-flight denitrogenation,compared with non-denitrogenated exposure to high altitude

1 hr 10,000 air 4 hrs 15,000 Control, noType of symptoms 3 hrs 15,000 02 on oxygen and denitrogenation

observed at 38,000 ft 3 hrs 38,000 3 hrs 38,000 3 hrs 38,000

Cumulated Cumulated Cumulatedn per cent n per cent n per cent

Completely symptom-free 22 76% 19 66% 11 27J%

Minor transient 10 pain 2 83% 5 83% 2 321

Mild 20 pain, regressive 3 93% 4 97% 2 3721

Moderate 30 pain,regressive 1 97% 0 97% 2 42J%

Non-regressive 30 pain 0 97% 0 97% 1 45%

Moderately severe 40 pain* 1 100% 1 100% 13 771%

Bends with VMR 0 -- 0 2 822

Mild or moderate chokes 0 -- 0 1 85%

Chokes with VMR 0 0 3 93

Chokes with bends and VMR 0 0 2 97C%

VMR and dizziness withoutother symptoms 0 0 1 100%

*Due to the enforcement of a conservative exposure policy 6n this project,cases of 40 pain were automatically removed from the chamber. One-third of thecontrol cases having 40 had progressed to 50 during the removal process. In thedenitrogenated tests, the individual with moderately severe pain in the "air plusoxygen" series was ordered down in accord with this practice, with momentary 50pain. He was not incapacitated, looked well, and probably could have "sat out"his bends since they exhibited a relatively flat-topped growth curve of the typethat frequently regresses. The corresponding individual in the series involving4 hrs at 15,000 ft was permitted to remain in the chamber, with some tendencytoward regression during the last hour at 38,000 ft.

WADC TR 53-227 27

RESTRICTED

RESTRICTED

Probability of Symptoms after Four Hours Denitrogenation

The prediction from the data is: Symptom free or minor 10 pain, 83.7%±5.7%,symptom free or no more than moderate 30 pain, 94.2-+3.6%; moderately severe painthat may or may not regress within 3 hours, 3.5%4+.8%; incapacitation, nodefinite figure possible but not to excede an occasional case per 50 to 100exposures (i.e. 1 or 2%).

These figures (divided by 100) represent the average individual probabilityof casualty. To apply thbm to a practical situation, it is necessary to con-sider the size of the air crew and the number of essential positions in the air-craft. Methods for doing this are available (5, PP. 11-12). For example, usingthe 1% casualty figure and considering all crew members essential, a three-placeaircraft would have a probability of 0.993 = 0.97 of completing the mission andthe abortion rate would be 100 - 97 = 3%. For a ten-place aircraft, the computa-tions would be 0.9910 = 0.90 probability of completion and 10% abortion. Assumingan individual casualty rate of K%, the abortion rate would be 1 for a 3 mancrew and 21% for a 10 man crew; assuming a 2% individual figure, the abortionrate would be 6% for a 3 man crew and 18% for a 10 man crew.

The data were secured at 38,000 ft, as our safety restrictions did notpermit prolonged exposure to any higher altitude. Figure 2 (Section I) offersa basis for estimating that at 40,000 ft the pains would have been about 10greater, so the expectation would be 5.8% ± 4% of moderately severe or severepain at that altitude with a somewhat increased probability of incapacitation.

The over-all picture of symptoms after denitrogenation displayed inTable 4, together with the details of the original protocols, leads to a not-entirely-imaginative visualization as follows: A considerable number ofindividuals in the series have no aeroembolism whatever. These men would haveno symptoms on this particular exposure even though the altitude was increasedconsiderably. Another and sizable group has aeroembolism in some degree; inmost of these cases it is at a sub-clinical level and would be revealed if thealtitude were to be increased sufficiently. The next most frequent type ofcase actually has clinical bends, usually late appearing and transient atthe exposure altitude used, but potentially more severe should the altitudebe increased. A few individuals actually have moderately severe aeroembolism,which would be severe at a higher altitude. With more denitrogenation beforethe high altitude exposure, this "curve" would be shifted to the left; the numberof clinically important cases as well as the sub-clinical cases would decrease.With less denitrogenation the curve would shift to the right; the limitingsituation would then be the un-tenitrogenated exposure, with a majority of thecases clinically severe; perhaps 5 or 10% would be free of aeroembolism andabout 20% would have clinically silent aeroembolism that would appear at ahigher altitude. This picture is dynamic; it is nbt at all a matter of indi-viduals being either definitely protected or unprotected by some magic criticaldosage of preoxygenation; there is an interplay of a number of physiologic andenvironmental factors whose influences are in general known and understood,but predictible only in terms of statistical averages.

It may be mentioned at this point that the heavy exercise denitrogenationseries showed a much higher incidence of severe or moderately severe symptoms,18.5%f compared with 3.5% in the light exercise series. Using a one-tailed "t"distribution, which is the appropriate statistical hypothesis in this particular

WAfC TR 53-227 28

RESTRICTED

RESTRICTED

case, the difference is definitely significant ("t" - 1.67 required, 1.85 ob-served). The fact that the observed symptoms of this type were not only morenumerous but also more severe adds further weight to the conclusion.

Control Group

The number and severity of symptoms in the control group may seem surprising-ly high. It is not; the total incidence is in fact in excellent agreement withthe first extensive standardized series of 38,000 ft exposures at this laboratoryduring the 1942-45 researches (6, p. 48, designated Series A*). That is theseries in which the physical activity was very similar to that of the presentexperiment. A total symptom incidence of 75.5% was observed, which may be comparedwith the present incidence of 72.5%. The difference is non-significant since "t"= 0.15. Compared at the level of 30 symptoms or worse, the figures are 55.3% and62.5%; "t" - 0.70 which is again non-significant. Direct comparison of incapacita-tion rates is not possible. It may be noted that the earlier series resulted in42.1% incapacitation while the current experiment resulted in 55.0* individualswith 40 pain or worse symptoms. If we make the reasonable assumption that onlyhalf of the cases of 40 pain would have become incapacitated, the agreement wouldbe almost exact. It would accordingly seem that the subjects and techniques ofthe current experiment have yielded representative results.

Validity of Preoxygenation Tables

In using the data of the present experiments to validate the Jones preoxygen-ation tables (3, P. 318; 8, p. 98), his factor of 90.7% protection or 9.3% remain-ing aeroembolism is employed as the probable protection coefficient for 4 hrs ofpreoxygenation. (In the table as printed, the figures have been rounded off to91 and 9%). Multiplying the symptom rates in the control by 9.3 we have the ex-pectation of 5.1% of 40 or worse symptoms, compared with the actual observationof 3.5 ± 2.8%4; 5.8% of 30 or worse symptoms compared with the observation of5.8 ± 3.6%. The agreement is very close indeed; the Jones tables are valid forsymptoms of this type.

The discrepancy is-considerable for symptoms of 20 intensity or worse (6.3%predicted vs 16.3% ± 5.7% observed), although it is possibly within the limitsof sampling error. However, when the minor 10 pains are included in the totalsymptom incidence, the prediction from the tables, 6.7%, is not very close to theobserved incidence of 29% after in-flight denitrogenation. There is a temptationto dismiss these 1 minor pains, which were reported by 13% of subjects who wereotherwise symptom-free, as being non-existent in most cases. It is quite truethat with an introspective set for pain observance, it might be expected thatafter having been in the altitude chamber for approximately six hours, some ofthe men would occasionally report minor joint pains that are in fact not bends.On the other hand, it has been a consistent observation in. other preoxygenationexperiments of shorter duration that the per cent reduction in moderate or severesymptoms is greater than for mild symptoms. In the Fraser stepwise ascent d4ta(2) for example, 1 hr on oxygen at 10,000 ft only reduced the mild symptoms 15%,although moderate and severe symptoms were reduced 63%; in his 20,000 ft experi-ment the mild symptoms increased although moderate and ,severe symptoms weregreatly reduced. In the three preoxygenation experiments reported by Henry andCook (7), the reduction in symptoms of 30 or worse was always greater than the

WLDC TH 53-227 29

RESTRICTED

RESTRICTED

reduction in total symptoms; indeed, one of the conclusions of that study wasthat the principle effect of the denitrogenation is to reduce the intensity ofsymptoms, rather than to delay their onset by an amount that is of practicalimportance.

It is entirely reasonable for preoxygenation to result in a greater propor-tion of minor symptoms than is observed in the controls-in fact, that is exactlywhat would be expected in the situation where there has been enough denitrogena-tion to almost but not quite prevent the occurrence of perceptible pain. Theuse of available denitrogenation tables must therefore be conditioned by therealization that they are valid for predicting the reduction in the serious typesof aeroembolism, but over-estimate considerably the reduction in total symptomincidence. This error may be expected to be cumulative; it probably increaseswith longer periods of denitrogenation. It may be noted that the use of suchtables is ordinarily concerned with severe symptoms.

Jones has also prepared tables for the in-flight denitrogenation situationwith air breathing at 10,000 ft followed by denitrogenation with pure oxygen(3, P. 320; 8, p. 100). It must be kept in mind that he had very little basicexperimental data available for these particular tables, and for some reason didnot take into account the increase in denitrogenation rate caused by mildhypoxia as mentioned earlier in Section III. It would accordingly be expectedthat these tables would under-estimate the protection. The data of the presentexperiment support this argument. For the air-oxygen profile the observed figuresare 3-5$ symptoms of 40 or worse compared with 7.7% predicted from his tablesand 6.9% symptoms of 30 or worse compared with 8.8% predicted. In the case ofminor symptoms the tables are not applicable; this matter has been discussed inpreceding paragraphs.

In-flight denitrogenation tables recently prepared by the writer (5, P. 19)do not require the "hypoxia" correction since there is presupposed at least 4 hrsof ambient air breathing at 10,000 ft which would very nearly reach the maximumpossible denitrogenation for this condition even with the slower rate used byJones. The use of this table may however be extended since the faster rate wouldachieve 90% of the maximum in li hrs and almost completely reach it in 2 hrs. Itshould be emphasized that this amount of denitrogenation (which is a valuablesupplement to the greater amount secured later by breathing pure oxygen) will belessened if the cabin altitude is lower than 10,000 ft. The experimentallydetermined incidence of serious symptoms after either of the flight profiles ofthe present study indicates that the protection is somewhat greater than predictedby the writer's tables. It is recommended however that the tables be used as abasis for planning, since it cannot be expected that the oxygen discipline willbe as carefully followed under operational conditions as has been the case inthe laboratory situation. As it stands the table is probably applicable withoutqualification to 40,000 ft exposures.

WADC ?R 53-227 30

RESTRICTED

RESTRICTED

DENITROGENATION AND SYMPTOM DELAY

Amount of Delay

It should be re-emphasized that denitrogenation, whether accomplished inflight or at ground level, has only a small influence on the time of symptomappearance (7, P. 355). The onset time of symptoms is correlated with theireventual severity; on the average, late-appearing symptoms tend to regress andare usually less severe than the early-appearing type (3, P. 334; 8, p. 158).With denitrogenation, these early and severe symptoms are converted to the laterand milder type, while the formerly mild and late have been so weakened that theyare sub-clinical. (The terms "early" and "late" are of course entirely relative,since symptom onset times occur in a continuous distribution). Data from thepresent experiments are in agreement with these concepts.

Table 5

ORIGINAL ONSET TIME OF SYMPTOMS

Arranged according to the eventual severity attainedduring the 7 hr flights. The beginning of justnoticeable pain is stated in minutes after reaching38,000 ft. (Entries in parenthesis give the time ofmaximum severity for the case listed directly above).

Maximum Pain Intensity

10 20 30 40 50

60 44 60 47 5073 58 (7o) (52) (88)

100 60 ......103 80 -- -....lo4 90 ......116 101 ......120 110 ......

In Table 5, it may be seen that after 4 hre of in-flight denitrogenation,the symptoms of possible practical importance have had an onset within the firsthour after reaching 38,000 ft and-have reached their maximum intensity within ahalf hour thereafter. Bends pain that never went higher than 20 began on theaverage about 20 minutes later than the initial onset of the more severe pains.The average onset time of the minor 10 pains was about 20 minutes later than forthe 20 pains. In these 58 tests, there was no instance of a new case of bendsappearing after 2 hre at j8,O00 ft. This means, of course, that the incidenceof aeroembolism does not accumulate hour by hour as the exposure time is lengthened.On the contrary, the first 2 hrs at high altitude represent the hazardous period.Occasionally, a severe case may occur later than ttia, but the evidence indicatesthat such a circumstance will be rare.

WADC M 53-227 31

RESTRICTED

RESTRICTED

It should be pofnted out that the data of Table 5 are completely at variance

with certain theoretical contentions that have been advanced by others (3,p. 255-257). If it were true that denitrogenation functions by delaying symptomappearance, the observed data could only be accounted for by postulating that the

reactors in the present group had unusually slow nitrogen elimination rates sincetheir delay was less than an hour, the non-reactors had very fast rates sincetheir delay was more than three hours, and there were no average men with middle-of-the-road rates of nitrogen elimination. This is entirely unreasonable. Thepresent writer is accordingly forced to reject the theory in favor of the Jonestheoretical explanations (3, p. 298 ff.; 8, p. 40 ff.) which do agree with theobserved facts of the current experiments as well as other published data,

In the control series, the average onset time of all symptoms was 49 min..In the denitrogenation series, the figure was 81 min. This onset delay of 32 minmay be compared with the Henry and Cook (7) report of 22 min delay in an oldergroup that was less well protected by 4 hrs of preoxygenation. In neither caseis the delay of impressive magnitude. Another way to calculate the delay is tocompare the onset times of the control and denitrogenated groups at equalpercentiles, which would in effect equate the individual nitrogen eliminationrates. (Since the intercorrelation of rates on two different days is by nomeans perfect, this method will bias the average in favor of a longer delay).This has been done with the data as arranged in Fig. 9. Computed in this manner,the average delay is 70 min when there is 4 hrs denitrogenation. The prophylacticvalue of denitrogenation must therefore be explained in terms of reduced symptomintensity rather than in delay of symptom occurrence.

Theoretical Time of Onset

A previous report (5, P. 5) gave several illustrations of the use of the

formula

dy/dt = ale-klt - a 2 e-k2t

to describe mathematically the rate of appearance of aeroembolism symptoms in a

group of individuals exposed to high altitude. Nims has used this type of formulato give a theoretical explanation of the physiology of bubble growth and decay,identifying k, with the diffusion constant governing the exchange of nitrogenbetween the tissues and alveolar air, and k with the diffusion constant for the

gas exchange between the tissues and the pain-causing bubble (3, p. 213). Simpli-fying his formula 41 by dropping the constant term, which may be done by using the

delayed onset time, 32 min, as ot (the point at which symptoms will first exceedthreshold magnitude and a1 will equal a 2 ), it is possible to describe mathematical-ly the rate of new cases of aeroembolism per 100 men per 5 min elapsed time. Theappropriate numerical values are a• = a 2 = 3.53; k' = 0.0133 and k2 = 0.077. The"half-time" rate coefficients (defined as 0.693/k) for these k's are 9 min for the

symptom growth rate and 52 min for the denitrogenation rate that controls thedecline of symptom incidence. The latter figure is reasonably typical for thisstage of denitrogenation (5, P. 16).

A curve of this function is shown in Fig. 8. The integral of the curve is

drawn as a smooth line in Fig. 9, where it is compared with the experimentallyobtained points giving the onset of symptoms during the denitrogenated high

altitude exposure. While no new cases actually appeared during the last hour

WADC M 53-227 32

RESTRICTED

RESTRICTED

160-

0o1 0020

TIME AT HIGH ALTITUDE (MIN) tT OS 4a F6SYMPTOM ONSET TIME JMIN)

Figure S. Rate of appearance of aero- Figure, . Cumulated incidenceembolism cases. (per 100 men per 5 min.) of aeroemboliem.after 4 hrs. of denitrogenation

of the 3 hr exposure, the theoretical curve predicts that a few new cases shouldhave appeared; there should have been 8.8 per 100 men, (i.e., 5 such cases in thepresent experiment). No more than 2/17 (i.e. 12%) of these 5 cases would havebeen expected to have symptoms as severe as 40 pain. In view of the tendency ofthe late appearing cases to be less severe, there is reason to believe that thisproportion ought to be even smaller. With a fourth hour of exposure, the curvepredicts 4 new cases per 100; probably none of these would be severe. It shouldbe mentioned that it was not possible to draw a theoretical curve for severesymptomb because the incidence of such cases was too small to furnish the neces-sary basic data.

As a matter of minor interest, the cumulative incidence of new cases in thecontrol series has also been plotted in Fig. 9. The two curves, control anddenitrogenated, both approach an asymptote. They appear superficially to be ofdifferent form, but this is simply a by-product of having plotted the data onthe usual probability ordinate, using however a logarithmic abscissa in orderto obtain even spacing of the points. (This statement can easily be verified byplotting the curves on linear coordinates). The formula is the same for bothcurves, except that for the controls, t is taken at 3 min. The curve constantsare of course different; for the controls, the intercept at t is higher(a1 - a2 8.25), the growth component is much faster (k 2 - 0?266), and thedenitrogenation rate is somewhat faster (kI S 0.0198) since the high altitudeexposure has occurred at an earlier position on the complete denitrogenationcurve (5, P. 16).

While the denitrogenation rate of 52 min half-time calculated from Fig. 9and shown as the descending limb of Fig. 8 is of representative magnitude, it is30 or 40% faster than would be observed at 30,000 ft or lower altitude (3, P. 309;8, p. 55). For the purpose of constructing protection tables, a representativerate for young men is 70 min half-time. The Jones tables used this figure for"probable protection". As the present experiment has confirmed the predictionsfrom his table, no modification appears to be indicated. Some further explanationof the practical use of his method will be given in the next section.

W-OC M 53-227 33

RESTRICTED

RESTRICTEDSECTION VI

RENITROGENATION SUBSEQUENT TO IN-FLIGHT DENITROGENATION

Loss of Protection

Interest in this topic stems from the practical question of how soon suscep-tibility returns when an individual with regressed aeroembolism (either manifestor silent) at high altitude, resumes ambient air breathing at low altitude or atground. Apparently this has never been investigated experimentally. An equallyimportant practical problem, related but not necessarily identical, concerns theestimation of protection loss when preoxygenation is interrupted by air breathing.Some experimental work on this problem has been reported (3, pp. 264 and 266),although most data have been secured after 8 or 9 hrs of denitrogenation followedby a 5 hr interruption, which unfortunately is not very helpful in a systematicexamination of a wide range of conditions.

Two theoretical treatments of the latter problem are available, by Bateman(3, p. 262 ff.) and by Jones (3, P- 313; 8, p. 91). The Jones graphic methodis emphasized here because it can easily be used for practical calculations.While it over-simplifies the theory of inert gas exchange by using only a singleexponential term instead of the four known to be involved (3, p. 293; 8, p. 83),it is only possible to quantify and validate a single component from clinicalaeroembolism data (3, p- 301; 8, p. 43). For this reason, both denitrogenationtables and renitrogenation tables are based on a one-term exponential system.The rate coefficient is most conveniently measured as half-time (defined as0.693/k, and easily determined graphically).

A typical example might be an individual with a denitrogenation half-timeof 70 min (Jones curve J). Having denitrogenated 4 hrs, an interruption of 1 hrwould place this individual back to the equivalent of 70 min of denitrogenation,i.e., the last 3 hrs (nearly) of denitrogenation would be lost by the interrup-tion. A 30 min interruption would lose about 2 hrs of denitrogenation; 15 minwould lose about l½ hrs and a 5 min interruption would negate the last 37 min ofdenitrogenation. An individual who received protection faster than the averagewould also lose his protection faster if renitrogenation occurred.

Method of Calculation

The writer has on several occasions been requested to explain (at thepractical level) the method of making such calculations. Figure 10 is a semi-log plot of per cent symptoms vs denitrogenation time. The straight-line curve,drawn with a 70 min half-time for the rate coefficient, permita the readingof per cent of original symptoms retained after any particular amount ofdenitrogenation time. Subtracting this figure from 100 gives the per cent protec-tion that appears in the Jones table, e.g. 91* after 4 hrs. Under the assumptionthat renitroglnation is the mirror image of denitrogenation, a second line is

rawn parallel to the first, with the intercept at 91 . Reading on this line,hr Interruption by breathing ambient air gives 67T protection retained;

100 - 67 = 33ý of original symptoms now present. Locating on the original linethe time required for 335 of symptoms retained, it is seen that 1 hr 51 mindenitrogenation reduces symptoms to this amount. Subtracting that time fromthe original 4 hrs shows that the last 2 hrs 9 min of denitrogenation has beenlost by the i hr interruption. Had the same interruption occurred at 3 hrsinstead of 4, the intercept of the auxilliary line would be at 100.0 - 16.6 - 83.4%.

TR 53-227 34

RESTRICTED

RESTRICTED

protection, there would be 6296 of protection retained (3O of original symptomsnow present), and the time loss would be 3 hrs minus 1 hr 37 min (i.e. the last1 hr 23 min of denitrogenation would be negated by the interruption and this ad-ditional time would be necessary to recapture the protection that had beenachieved by the original 3 frs).

While the same general principles might be expected to apply to the in-flightdenitrogenation-renitrogenation problem, it is necessary to use rate coefficientsknown to be applicable to the si.tuations under consideration. This modificationwill be discussed after presenting some illustrative experimental results.

0 1 )XITROGMUTION TIN& (HRS)

90 I 2 4

l0l

1/2?

15-

= II

z

0900

V FA NI TN T S050 X-10,000 FT. 70

.0.-5,000 FT. so-0 VIPAIS I

DENITROGEHATION TIME (HftS)

Figure 10. Denitrogenation - Renitrogen-ation calculations. Insert graphs show datafrom storage experiments.

Experimental Results

Eight men who were in the 10,000 ft "storage" experiment were persuaded toremain available for a 38,000 ft exposure after renitrogenating by breathingambient air at ground level. There was pre-established bends of 40 intensityat the beginning of the 10,000 ft storage period. Pure oxygen was breathed dur-ing storage. Each man had remained in storage until his symptoms (silent at10,000 ft but manifest during the "tests" at higher altitude) regressed to thepoint that they did not show at 40,000 ft. Presumably this regression was causedby in-flight denitrogenation. In the final 38,000 ft exposure after renitrogena-tion the standard step-up exercise was performed at 21 min intervals.

Two cases, Nos. 189 and 191, were permitted to renitrogenate for 25 min andthen taken to 38,000 ft. While no bends occurred, these two runs are not usable

WK&c Tn 53-227 35

RESTRICTED

RESTRICTED

because abdominal gas pain forced descent after a few minutes exposure. Theother 6 cases (with longer renitrogenation periods) are usable. Table 6 sum-marizes the details. Length of the renitrogenation period is assumed to be theelapsed time from oxygen off during descent from storage to oxygen on at 12,000ft during ascent for the final 38,000 ft exposure.

To estimate the in-flight denitrogenation rate coefficient, the data fromTable 3 (Section III) have been used to plot the points of the insert graphs ofFig. 10. This procedure assumes that the per cent of individuals reacting withbends pain after varying lengths of storage is a measure of the bends-producingnitrogen gas remaining in the group of stored individuals; the rate of reductionin the proportion reacting is therefore a measure of the average rate of in-flightdenitrogenation. The 10,000 and 15,000 ft series have both been plotted, sinceother analyses indicate the results are similar in both. The average half-timedenitrogenation rate coefficient from these data, estimated from the straightlines of the Fig. 10 insert graphs, is 100 min. While slightly slower thanaverage rate constants obtained by others with different methods, using ground-level denitrogenation (3, P. 313; 8, p. 91), it is remarkably similar to thosedeterminations. Since the storage series necessarily involved some selection,as it consists entirely of individuals who developed bends in their first 38,000ft exposure, it would be expected to yield a somewhat slower denitrogenationrate than an unselected group.

Using the method previously explained, with the substitution of theindividual denitrogenation and renitrogenation times of Table 6 and the 100 minhalf-time coefficient, we would expect 61% recurrence for the first two men,4%8 for the middle two, and 91 and 88% for the last two men who had over 4 hrs

renitrogenation. Assuming a 70 min half-time, these figures would be 60% forthe first two, 52$ for the second two, and 96 and 94% for the last two men.These calculations assume 100% susceptibility, since each man originally had 40bends. This may be slightly high, but on the other hand even an unselected groupwould have a very high symptom incidence with the altitude and exercise used,so the discrepancy cannot be large. Inspection of the 38,000 ft results ofTable 6, in relation to the above computations, shows that they are in line withtheoretical expectations.

WADC TR 53-227 36

RESTRICTED

RESTRICTED

Table 6

DENITROGENATION-RENITROGENATION PROTOCOLS

Bends of 40 intensity had been pre-established at 38,000 ft just beforethe in-flight denitrogenation started at 10,000 ft breathing pure oxygen.

Subject Length of in-flight Ambient air Symptoms at 38,000 ftNumber denitrogenation at ground after renitrogenation

197 2 hrs 4 min 56 min No symptoms; 40 min exposure.

206 2 hrs 5 min 58 min 1° knee at 9 min, 30 at 12 min;50 at 15 min; descent.

240 3 hrs 39 min 56 min Intermittant 10 ankle after34 min. Exposure 70 ain.

223 3 hrs 33 min 57 min 10 ankle after 30 min;"fluctuated 00- 20 for 21 min, andthen regressed. Exposure 70 min.

200 3 hrs 21 min 4 hra 10 shoulder 31 min; reached55 min 40 at 33 min; descent.

(explosive type symptoms)

202 3 hrs 35 min 4 hrs 10 knee after 16 min; 30 at30 min 19 min; descent at 24 min.

SUMMARY

When aviator t s "bends" pain was pre-established by exposure to a decompres-sion chamber pressure equivalent to 38,000 ft, the average critical altitudefor disappearance or reappearance of mild pain was 23,900 ft. For moderate pain,the critical altitude was 28,500 ft; for severe pain, it was 31,600 ft. Thestandard deviation averaged about 20% of the mean. Tables were prepared to showthe range of individual differences in critical altitude for the variousintensities of pain.

Rapidly repeated ascent and descent of several thousand feet in the regionof the critical altitude for bends did not cause any appreciable alteration inaeroembolism symptoms as compared with steady altitude conditions. It wasobserved that a symptom growth phase occurred during the first 40 min, followedby a symptom decay phase that was progressive. These phases occurred in bothvariable altitude and steady altitude conditions.

Men with pre-established bends were "stored" without exercise at altitudesranging from 10,000 to 25,000 ft. When stored at 10,000 or 15,000 ft, the

WA=O TR 53-227 37

RESTRICTED

RESTRICTED

critical altitude increased progressively with time, resulting in a gain of12,000 or 13,000 ft in the course of 3 hrs. The percentage of individuals whowere bends-free at the maximum test altitude of 40,000 ft also increased progres-sively, in agreement with a logarithmic law that is characteristic of the knownrate of denitrogenation as estimated by other methods and in other circumstances.On the other hand, men stored at 20,000 or 25,000 ft showed relatively smallgain in critical altitude during 3 hrs. Observation of a decline in criticalaltitude during the first i hr of storage suggested that "silent bubble" growthat these altitudes might be the causal agent. These results were considered tooffer a basis for concluding that in-flight denitrogenation prophylaxis prior-tohigh altitude exposure could be effectively accomplished at 10,000-15,000 ft butnot at 20,000-25,000 ft.

Comparison of results at 10,000 ft breathing ambient (cabin) air vs breathingpure oxygen showed that the amount of denitrogenation, as estimated from thechange in critical pain altitude, was the same for both conditions during thefirst 60 to 90 min of storage. Presumably the increase in denitrogenation ratecaused by the mild hypoxia when breathing air was as important a factor as thegreater nitrogen differential when breathing pure oxygen. After 1 or I* hrs time,there was relatively little change in critical altitude during air-breathingwhereas it continued to increase during oxygen breathing. The similarity betweenthe two 10,000 ft series during the first hour was not due to the recompressionfactor, since the increase in altitude was not different from that observed inthe 15,000 ft series. These results were interpreted as suggesting that in-flightdenitrogenation prophylaxis of several hours duration should be almost fullyeffective if ambient air instead of oxygen was bireathed at 10,000 ft during thefirst hour. This procedure would result in considerable economy of oxygen andlessened discomfort to flight personnel. The desirability of increasing the cabinaltitude to 12,000 or 14,000 ft during the last half hour of air breathing beforeshifting to pure oxygen was pointed out.

In order to validate existing tables for estimating the amount of denitro-genation via preoxygenation required to prevent severe bends and other dangeroussymptoms of aeroembolism, and to confirm directly the conclusions of the precedingparagraph, two types of low altitude in-flight denitrogenation profiles werecompared as to their effectiveness in preventing aeroembolism at high altitude."One hour on ambient (cabin) air at 10,000 ft was as effective as 1 hr on pureoxygen at 15,000 ft, when both were followed by 3 hrs on oxygen at 15,000 ft and3 hrs of test exposure to 38,000 ft with light physical activity. This amountof denitrogenation resulted in the amount of protection against serious symptomsthat was predicted for 4 hrs of oxygen breathing by available preoxygenationtables. Under control conditions (without denitrogenation before exposure to38,000 ft), the incidence of moderately severe or worse symptoms was 55%; after4 hrs of in-flight denitrogenation the figure was only 3.5%, and there were nocases of chokes, vasomotor collapse or fulminating severe bends although suchsymptoms were fairly common in the control series. Mild symptoms showed lessthan the predicted reduction.

Symptom onset at high altitude was delayed 32 min (or 70 min, depending onthe method of calculation) as a result of 4 hrs denitrogenation. The observedfacts were inconsistent with the hypothesis that the prophylaxis from denitrogena-tion could be explained by delayed symptom onset. On the contrary, the chiefeffect was a decreased symptom intensity. The rate of occurrence of symptomsreached its peak within an hour after reaching high altitude and declined rapidly

WADC M 53-227 38

RESTRICTED

RESTRICTED

thereafter at the rate of 52 min half-time, which is a typical denitrogenationrate for 38,000 ft.

With relatively heavy physical exercise during the denitrogenation periodat 15,000 ft, silent bubble formation occurred in several cases. When this typeof exercise was engaged in at 38,000 ft, it was found that 4 hrs of denitrogena-tion failed to offer adequate protection against aeroembolism, since 18.5% ofthe men had chokes, severe bends or related vasomotor reactions. A considerationof the theoretical factors involved (carbon dioxide production in particular)suggested that adequate protection would be difficult to achieve under conditionsof heavy exercise.

The in-flight denitrogenation rate estimated from symptomatic evidence ofsilent bubble decay was observed to have a half-time coefficient of 100 min(somewhat slower than typical rates observed by others at ground level). Re-nitrogenation at zero altitude, breathing ambient air, apparently progressed asthe mirror image of denitrogenation (both being estimated on the basis ofaeroembolism occurrence). Individuals who had denitrogenated to the point ofsymptom disappearance in flight, recovered approximately 50% of original suscep-tibility in an hour of air breathing and 90% in about 41 hrs, as predicted bytheory.

Symptom incidence in both the control series and the unselected group usedto obtain pre-established bends in the storage experiments, compared with theincidence reported some ten years earlier for similar high altitude exposures,indicated good reproducibility and suggested that the individuals used in thepresent study constituted a representative group of young men. The necessityfor standardization of method with consideration of psychological factors wasemphasized. In general, the experimental findings in the study were consistentwith aeroembolism and denitrogenation theory developed in previous reports fromthis laboratory.

WAN N 53-227 39

RESTRICTED

RESTRICTED

BIBLIOGRAPHY

1. Alder, H.F. Neurocirculatory Collapse Resulting From Exposure to SimulatedHigh Altitude in the Decompression Chamber. United States Air Force SpecialReport. United States Air Force, School of Aviation Medicine, Randolph AirForce Base, Texas, June 1950.

2. Fraser, A.M. Effect of Stepwise Ascent on the Incidence of DecompressionSickness. Report No. C-2436. Proc. C.A.M.R., National Research Council,Ottawa, Canada, February 1943.

3. Fulton, J.F. (ed.). Decompression Sickness. W.B. Saunders Company,Philadelphia, 1951.

4. Gray, J.S. Effect of Denitrogenation at Various Altitudes on Aeroembolismin Cadets. Report No. 258, C.A.M., National Research Council, Washington, D. C.,J anu ary 1944.

5. Henry, F.M. The Aeroembolism Problem for Long-Range Missions. WADC TechnicalReport No. 52- Wright Air Developme-nt Center, February 1952 (RESTRICTEDreport, titJe UNCLASSIFIED).

6. Henry, F.M. Altitude Pain: A Study of Individual Differences in Suscep-tibility to Bends, Chokes, and Related Symptoms. Journal of AviationMedicine, Volume 17, February 1946, pp. 28-55.

7. Henry, F.M. and S.F. Cook. Effectiveness of Pre-flight Oxygen Breathing inPreventing Decompression Sickness. Journal of Aviation Medicine, Volume 16,October 1945, pp. 350-355.

8. Lawrence, J.H., et al. Studies on Gas Exchange. United States Air ForceMemorandum Report No. MCREXD-696-l-, United States Air Force, Air MaterielCommand, March 1948.

9. Rodbard, S. Recurrence of Decompression Sickness on Reascent to HiAltitudes. Air Surgeont s Bulletin, Volume 1, No. 11, November 19J4,pp. 6-7.

10. Stewart, C.B. and H.W. Smith. Effect of Reascent on Recurrence of Decompres-sion Sickness. Report No. C-2691, Proc. C.A.M.R., National Research Council,Ottawa, Canada, October 1943.

AR MR 53-227 T40

RESTRICTED

RESTRICTED

APPENDIX I

SYMPTOMS, OTHER THAN BENDS, RESULTING FROM

ALTITUDE CHAMBER DECOMPRESSION

Donald J. Rosenthal, M.D.

University of California, Berkeley

ABSTRACT

1. The most common symptoms resulting from decompression (except for jointpains) were abdominal gas pains, aerotalgia, chokes, syncopal reactions, andscotomata, in that order of frequency. Only one mottled skin lesion was noted.

2. No really serious reactions were produced, and there were no delayedcollapse reactions or paralyses. The absence of dangerous complications inthese experiments is thought to be due to a policy of rapid removal of subjectsfron the chamber at the first siqn of the development of any of the morehazardous reactions.

WDC TR 53-227 41

RESTRICTED

RESTRICTED

The most common symptoms resulting from decompression (except for jointpains) were abdominal gas pains, aerotalgia, chokes, syncopal reactions, andscotomata, in that order of frequency. Only one mottled skin lesion was noted.

No really serious reactions were produced, and there were no delayedcollapse reactions or paralyses. The absence of dangerous complications inthese experiments is thought to be due to a policy of rapid removal of subjectsfrom the chamber at the first sign of the development of any of the morehazardous reactions.

The symptoms which may appear in an individual exposed to reducedatmospheric pressure have already been mentioned briefly in the introductionof this report. Several excellent reviews of these phenomena have appearedin the literature in the past ten years. (1,2,3). Interest has revived inthis field recently, spurred on by the development of aircraft which are ableto fly higher and higher with each new model produced. It is therefore thoughtdesirable to add to the report a short discussion of the symptoms, other than"bends", which were observed in the experimental subjects while they were atreduced pressures in the altitude chamber.

All of the subjects were male university students, 88% falling in the agegroup 21-25 years. Prior to their first experimental flight, they were givena medical survey, with emphasis placed on the condition of their ears, noseand throat, lungs, and heart. Any pathology of these structures was sufficientto exclude an individual from the study. In addition, a history of a recentrespiratory or sinus infection, recent allergic rhinitis or asthma, and migraineor a convulsive disorder at any time in the past would also disqualify apotential subject from the group.

Following the physical examination and indoctrination lecture, the menwere given a short flight to a simulated altitude of 30,000 ft, to familiarizethem with the oxygen equipment and with the techniques useful in equalizing thepressure in the middle ear. The descent from this altitude was made at anaverage rate of 4,500 ft per min. All subjects who experienced great difficultyin clearing their ears were eliminated from future flights.

It should also be mentioned here that it was agreed, at the onset of theexperimental work, to take a conservative attitude towards the development ofsymptoms other than joint pains or gas pains, and it was decided to limit thesetwo manifestations of the reduced environmental pressure to a severity of 40 or50 (California Altitude Pain Scale). While subjects were in the chamber underreduced pressure, they were under the continuous scrutiny of one inside observerand two or more observers stationed outside the chamber at the observationports. All of these observers were aware of the nature, manifestations, andtreatment of the various reactions which may occur at altitude. When any ofthem felt that a subject was beginning to display signs of an untoward reaction,the subject was removed from the chamber without delay. Discussions concerningthe exact diagnosis of the symptoms presenting themselves, and evaluation oftheir actual severity, were held afterwards, with the subject out of the chamber.

It is felt that the complete absence of severe reactions, and the relativepaucity of all types of untoward reactions, resulted from the pre-flight screeningand orientation and the conservative policy described in the above paragraph.

WLDC M 53-227 42

RESTRICTED

RESTRICTED

Chest Symptoms

Of a total of 375 man runs in the chamber, 23 cases of "chokes" (3,4)were seen (Table 7). Fourteen of these cases appeared during the storage runs,two during the cycles experiments, one during the heavy exercise series, andsix during the in-flight denitrogenation control flights (Table 7). As hasbeen reported by nearly all the observers in the past, there was a significantassociation between the chest symptoms and joint pains. Of the 23 cases, 13were classified as mild or incipient chokes, in accordance with Bridgets classi-fication (3), and 10 were rated as moderate in severity. No severe cases wereseen, probably due to the rapid removal from the chamber of any subject whocomplained of a 20 or greater chest pain, or who was observed to be coughingfrequently.

With regard to the time of onset of chest symptoms, there appears to be adefinite relationship with physical exertion, as is the case with bends. Ofthe 16 cases seen during the storage and cycles runs, where the subject exercisedat 38,000 feet in order to establish bends pain for study during the experiments,11 cases of chokes appeared in 30 min time or less. Of the six cases seen duringthe in-flight denitrogenation controls, where there was no regular exercise andthe entire flight was made at 38,000 feet, only two cases developed within 30min, and the other four developed later.

In only a few of these cases did the symptoms become worse during thedescent. Nearly all complained of a raw or tight feeling substernally, andthree men volunteered the description that the chest sensation was similarto that experienced after running a hard, long race. Deep inspiration madethe pain worse in every case.

There were no auscultatory findings present at any time in these men, andall felt well by the time they were removed from the chamber. The only residualthen, in several of the cases, was a mild. uncontrollable cough on deep inspira-tion, and in every instance, this had ceased 30 min after the individual hadreached ground pressure. Cyanosis was seen in two cases, and it disappearedbefore the cough did, in both subjects.

Vasomotor Reactions

These reactions have been described and discussed in great detail elsewhere(1,2,3) and their general character will not be set forth here.. As far as thecases seen in this series of experiments are concerned, several were probablydue to hyperventilation, secondary to anxiety or to severe joint or gas pains.However, this factor is a difficult one to evaluate, particularly in the experi-mental situation obtaining. There is no doubt that the psychic effect of severepain is, in some individuals, sufficient to cause a syncopal reaction without theadded etiological factor of a respiratory alkalosis.

Only 19 instances of vasomotor reactions were seen in the entire series ofaltitude chamber exposures, and no actual case of syncope occurred. This latterfact is again probably explained by the prompt removal from the chamber of sub-jects evidencing premonitory symptoms. Ten of the reactions were seen during thestorage-type flights, eight during the in-flight denitrogenation controls, andone during the heavy-exercise series.

VAXO M 53-27 243

RESTRICTED

RESTRICTED

Table 7

SU1AWRY OF FINDINGS IN SUBJECTS EXHIBITING CHEST SYMPTOMS

Time •fBendsonset Severity2(min) Chest Pain Cough Mild Mod. Severe VMR

Storage Series

5 Mod. Yes - Yes - -

10 Mild .. ...10 Mod. - - Yes - -

23 Mild - Yes - - -

29 Mild Yes . ...30 Mod. - . ...30 Mod. - . ...30 Mild - Yes - - -

30 Mild Yes - Yes -

45 Mod. Yes -. ..

65 Mild Yes - Yes -

75 Mild - - - Yes95 Mild - - Yes -

120 Mild4 - - - Yes

"Cycles" Series

7 Mild Yes Yes J -20 Mild -- Yes-

Heavy Exercise Series

2'43 Mod. I ____I___________ IIn-Flight Denitrogenation Control Series

30 Mod. Yes . ...31 Mod. Yes - - - Yes15 Mild - - - Yes -

60 Mod. Yes Yes - - -

130 Mild Yes . ...160 Mod. Yes - Yes - Yes

1 After arrival at simulated altitude of 38,000 ft.2 Vasomotor reaction.3 Preceded by 4 hours at 15,000 ft on oxygen. Exercises every 2j min

after arriving at 38,000 ft.14 With visual symptoms.

WADeOT 53-227

RESTRICTED

RESTRICTED

As one might expect from the above discussion, the incidence of thesereactions was highest at the time when the bends pain was the greatest. Inthe storage group, where bends pains were induced by exercising for the first30 min at 38,000 ft, eight of the ten vasomotor reactions occurred within thefirst 30 min. In the in-flight denitrogenation control group, without regularexercise at 38,000 ft, only two of the eight cases developed within 30 mn.

Even more than in the case of chokes, pallor was found to be the bestearly sign of an impending vasomotor reaction (Table 8). It preceded subjectivecomplaints by a variable period of time, ranging from a minute or two to 10-15min. No cyanosis was observed, and aince the altitude was never over 40,000 ft,and the fit and functioning of the oxygen apparatus was obvious at all timesduring the flights (due to the use of rebreather bags and BLB masks), it isunlikely that hypOxia ever was an etiological factor in the reactions seenhere. It is of interest that in the in-flight denitrogenation group, kept at10,000 ft without supplementary oxygen for the first hour of the flight,cyanosis was seen in several of the individuals, and yet no vasomotor reactionsoccurred during that portion of the flight.

Table 8

INCIDENCE OF RELATED SYMPTOMATOLOGY IN 19 CASES OFVASOMOTOR REACTIONS

Per Cent ofSymptom Incidence 19 Cases

Pallor 14 74Nausea 11 58Diaphoresis 10 53Dizziness 9 47Faintness 9 47Bends-mild 5 26Bends-moderate 3 16Bends-severe 5 26"Chokes"-mild 3 16"Chokes"-moderate 1 5Abdominal Pain 3 16Visual Disturb. 1 5Headache 1 5

The tyoical clinical picture presented was one where the subject evidencedmarked pallor, with a drenching cold sweat. and complaints, usually of a markednausea, dizziness, weakness and faintness. In one case, headache was alsopresent. Simply lowering the head well below the knees, with the subject inthe sitting position (making sure, the meanwhile, that the oxygen mask remainedin place) was sufficient to alleviate the symptoms in all the cases, save one,before the subject had reached a pressure altitude of 10,000 ft, on his waydown and out of the chamber. A slight bradycardia (50-60/min) was usuallypresent when the symptoms were at their peak, and immediately after removal fromthe chamber, the blood pressure was normal in every case, except the one

WADC TR 53-227

RESTRICTED

RESTRICTED

mentioned above. In his case, it returned to normal levels, when the subject wasin the supine and erect positions, within 40 min.

These particular subjects were held at the chamber site, and observed fortwo hours following recovery from their reaction, and then allowed to leave.Telephonic inquiries as to their status were made during the next 12 hours. Allof the subjects in the entire experimental group were instructed to check with thephysician 12 hours after their flight, and sooner if symptoms required it. Nodelayed collapse reactions were seen.

Abdominal Pains

Other than noting the rather high incidence of discomfort due to gas painsat 38,000 ft, little mention will be made of this symptom. Only a small numberof runs were aborted due to this cause. Occasionally, when the subject was havinggreat difficulty in passing gas, it was found helpful to partially compress him,and urge him to massage his abdomen and actively try to pass flatus. On manyoccasions, this procedure enabled the subject to rid himself of the troublesomeintestinal gas, whereas previously, at the higher altitude, he had been unableto do so.Aerotalgia and Aerotitis

Due, probably, to the screening and indoctrination procedures, relativelyfew individuals developed otalgia during the actual test flights. Use of avasoconstrictor inhaler or nasal spray helped most of the individuals who haddifficulty in clearing their ears by any of the usual methods (swallowing,yawning, grimacing, Valsalva maneuver). The spray, which was of 0.05% naphazolinehydrochloride (Privine HCl) was much more effective than the inhaler.

In spite of these measures, several subjects developed moderately severeaerotitis, but all of these cases subsided within 48 hours with vasoconstrictortherapy alone, and none became secondarily infected. No ruptures of the tympanicmembrane were produced.

One individual had the rather harrowing experience of being "trapped" at asimulated altitude of 23,000 ft with severe bends pains in both knees, and severepains in both ears. Increasing the pressure, to relieve the joint pains, producedexcruciating ear pain, and decreasing the pressure, to relieve the ear pains,caused the bends to get much worse. He was sensitive to pressure changes of aslittle as 500 ft either way, at that altitude. By means of several applications ofthe nasal spray, and a rate of descent of 200 feet per minute, he was eventuallyremoved from the chamber without laceration of his ear drums.

Three cases of frontal sinus pain appeared during descents, all alleviated

by stopping the descent temporarily and using the nasal spray. One slight nasalhemorrhage was also noted.

WADC TR 53-227 46

RESTRICTED

RESTRICTED

Visual Disturbances

The only neurologic abnormality which appeared in this series of low pressureexposures was the presence, in a small number of subjects, of scotomata and theheadache which usually followed them (1, 5). In 375 man-runs, six cases developed.Five individuals were suffering from bends at the time their visual symptomsappeared. One of the cases developed five minutes after the individual had beenremoved from the chamber because of mild chokes. The other five all developedwhile the subjects were in the chamber, although in only two cases were the menat the highest altitude to which they had been exposed. The other three were attheir storage altitudes (20,000 ft in one case, 10,000 ft in the remaining two).

The scotomata were discovered in one individual after he complained that hewas unable to read; the other five complained of seeing flashes of light or brightspots before their eyes. The scotomata were variously characterized as beingscintillating, glistening, or shimmering (Table 9). All of them were multiple,

Table 9

SUMMARY OF CASES WITH VISUAL DISTURBANCES

Description of AssociatedScotomata Symptoms TimeI Headache

Bilateral, multiple, Bends 30 min Bilateral, diffuse,"shimmering" lasted 2 hours.

Bilateral, multiple, Bends & 7 min Bilateral, diffuserevolving, "bright" Chokes duration unknown.

Bilateral, multiple, Bends 31 hrs "Sharp", in rt."shimmering." Had occipital region,left homonymous started 1½ hrs afterhemianopsia scotomata appeared.

Lasted 6 hours.

Bilateral, multiple, None 30 min Bilateral, frontal."flashing" Started 30 minutes

after reaching ground.

Lasted 3 hours.

Bilateral, multiple Bends 10 min None.

Bilateral, multiple, Bends Desc. 2 Moderately severe left"scintillating" frontal, after 30 minutes

at ground. Accompaniedby nausea. Subject hadmigraine previously.Lasted six hours.

1 Time, in minutes or hours after reaching ground level, for

disappearance of scotomata.

2 Disappeared during descent.

WADO TR 53-227 47RESTRICTED

RESTRICTED

and homonymous, and they were all said to be moving in various ways. Visual fielddefects were demonstrable by confrontation tests in five of the six cases. Fiveof the cases were followed by headache. One subject, who had an almost completeleft homonymous hemianopsia, began to complain, 1½ hrs after he was out of thechamber, of a severe right occipital headache. The other headaches were more vaguein localization, but usually in the frontal region.

There were no muscular paralyses noted during the duration of the project,nor did any of the aphasias appear.

Skin Lesions

Only one significant skin lesion was noted--a mottled skin lesion (1).Many of the subjects noted chilly sensations and itching of the skin, but no recordwas kept of these findings since they were, in every case, trivial and not trouble-some.

The aforementioned subject had severe bends in his left shoulder during a 3 hrstorage run. The next morning, he complained of slight pain and swelling in theregion of the left pectoralis major muscle. The pain and swelling increasedthroughout that day, and were accompanied, later on, by a diffuse erythema andincreased warmth of the skin overlying the edematous and tender area. No indurationor crepitus was present. Forty-eight hours after the flight, this mottled skinlesion began to regress, and 24 hrs later, had almost completely disappeared.

REFERENCES

1. Ferris, E.B. & Engel, G.L.: The Clinical Nature of High AltitudeDecompression Sickness, Chap. II in Decompression Sickness, editedby Fulton, J.F., W.B. Saunders Company, Philadelphia, 1951.

2. Adler, H.F.: Neurocirculatory Collapse at Altitude. Special projectreport, USAF School of Aviation Medicine, Randolph AFB, June 1950.

3. Bridge, E.V., Henry, F.M., Cook, S.F., Williams, O.L., Lyons, W.R., andLawrence, J.H.: "Decompression sickness: Nature and incidence ofsymptoms during and after decompression to 38,000 feet for ninetyminutes with exercise during exposure." J. Aviat. Med., a5:316-327,1944.

4. Bridge, E.V., Henry, F.M., Williams, O.L., and Lawrence, J.H.; 'OChokes':A respiratory manifestation of aeroembolism in high altitude flying,"Ann. of Int. Med. 22:398-407, 1945.

5- Whitten, R.H.: "Scotoma as a complication of decompression sickness,"Arch. Ophthalmology, 36:220-224, 1946.

WL MB 53-227 48

RESTRICTED